Douwe Maan was GitLab’s 10th employee and is now the founder and CEO of Meltano, GitLab’s spin-off data operations platform. His story is fascinating, and his passion for an open source approach to data ops is contagious! Enjoy this great conversation, and please share with your network!
Watch this episode on YouTube: https://youtu.be/XhnJAA_6wrc
Your next business card will be a YouTube channel! – Learn more at https://algmin.com
Save 20% on your first order at the DATAVERSITY Training Center with promo code “AlgminDL” – https://training.dataversity.net/?utm_source=algmindl_res
About Douwe Maan:
Douwe Maan is the founder and CEO of Meltano, an open source data operations platform. Born and raised in the Netherlands, Douwe has been around computers since a young age. While a college student, he began working at GitLab as their tenth employee. While working for GitLab, Douwe spent six months travelling the world, visiting colleagues in 14 countries on 5 continents.
In late 2019, Douwe joined the internal Meltano project at GitLab and quickly became its General Manager. In early 2021, Douwe led Meltano in spinning out of GitLab to become an independent startup while raising $4.2 million in seed funding, led by Alphabet’s GV. Today, Meltano is on its way to becoming the world’s first open source data operations platform.
Meltano Home: https://meltano.com
Meltano Slack Community: https://meltano.com/slack
Douwe Maan on Twitter: https://twitter.com/DouweM
Episode Transcript:
1
00:00:00,000 –> 00:00:00,625
douwe_maan: So I joined GitLab as employee number
2
00:00:00,875 –> 00:00:04,291
douwe_maan: ten. No idea where it could go. And for those who haven’t seen the news,
3
00:00:04,375 –> 00:00:08,375
douwe_maan: GetLab actually went public on the NASDAQ Stock Exchange in October, So it’s
4
00:00:08,458 –> 00:00:12,250
douwe_maan: been an incredible journey that I had the privilege of seeing from the front
5
00:00:12,458 –> 00:00:14,625
douwe_maan: lines from this very early stage.
6
00:00:19,625 –> 00:00:22,958
anthony_algmin: Welcome to the Data Leadership Lessons podcast. I’m your host, Anthony J.
7
00:00:23,208 –> 00:00:26,541
anthony_algmin: Algmin. Data is everywhere in our businesses, and it takes leadership to make
8
00:00:26,625 –> 00:00:29,916
anthony_algmin: the most of it. We bring you the people stories and lessons to help you
9
00:00:30,083 –> 00:00:33,750
anthony_algmin: become a data leader. Our show is produced by Algmin Business Media, where we
10
00:00:33,833 –> 00:00:37,291
anthony_algmin: make having your own video podcast as easy as joining a call and sending an
11
00:00:37,333 –> 00:00:41,666
anthony_algmin: email. The stage is yours. Visit Algmin.com to learn more. Today on
12
00:00:42,000 –> 00:00:43,541
anthony_algmin: again to day under Italyship to day on data leadership lessons. We welcome
13
00:00:43,625 –> 00:00:44,083
anthony_algmin: Douwe Maan. Douwe is the founder and C e O of Meltano, an open source data
14
00:00:44,083 –> 00:00:44,666
anthony_algmin: Douwe Maan. Douwe is the founder and CEO of Meltano, an open source data
15
00:00:44,666 –> 00:00:46,000
anthony_algmin: Douwe Maan. Douwe is the founder and CEO of Meltano, an open source data
16
00:00:46,000 –> 00:00:48,541
anthony_algmin: Douwe Maan. Douwe is the founder and CEO of Meltano, an open source data
17
00:00:48,750 –> 00:00:50,458
anthony_algmin: operations platform.
18
00:00:50,458 –> 00:00:51,750
anthony_algmin: While a college student he began working at
19
00:00:51,833 –> 00:00:56,250
anthony_algmin: Gitlab as their tenth employee. In late twenty nineteen, Doa joined the
20
00:00:56,458 –> 00:01:00,708
anthony_algmin: internal Meltano project at Git Lab and quickly became its general manager
21
00:01:01,041 –> 00:01:05,375
anthony_algmin: and early twenty twenty one. Dowa led Malano and spinning out of Git Lab to
22
00:01:05,416 –> 00:01:09,041
anthony_algmin: become an independent Startu while raising four point two million dollars in
23
00:01:09,125 –> 00:01:12,625
anthony_algmin: seed funding to day. Meltano is on its way to becoming the world’s first
24
00:01:12,791 –> 00:01:16,333
anthony_algmin: open source data operations platform. Dowa welcome to the show,
25
00:01:16,916 –> 00:01:19,166
douwe_maan: thanks. Anthony. Uh, very excited. beer.
26
00:01:19,750 –> 00:01:22,583
anthony_algmin: So like we do with all our first time guests, Just take a moment. Give us a
27
00:01:22,583 –> 00:01:25,791
anthony_algmin: little bit more of the context of your career, kind of where you came from
28
00:01:25,958 –> 00:01:28,916
anthony_algmin: and how what you studied in school and how it led to this amazing
29
00:01:29,083 –> 00:01:33,458
anthony_algmin: opportunity to work with Meltano work with Giitlab. And you’ kind of be ond
30
00:01:33,541 –> 00:01:35,250
anthony_algmin: this journey that you on. It’s amazing
31
00:01:35,500 –> 00:01:39,666
douwe_maan: Yeah, my pleasure. Um, you already hit a couple of the points in your your brief
32
00:01:39,833 –> 00:01:43,833
douwe_maan: summary, Uh, which I’m I know you were reading from my bio. Uh, but the story
33
00:01:44,166 –> 00:01:47,916
douwe_maan: really starts pretty at this point eighteen years ago when I was nine years old
34
00:01:48,166 –> 00:01:51,375
douwe_maan: and I got into programming because a friend of mine showed me that there were
35
00:01:51,416 –> 00:01:54,541
douwe_maan: some websites that explained how you could write code and would make things
36
00:01:54,791 –> 00:01:58,250
douwe_maan: appear in your web browser, which of course felt incredibly powerful. I still
37
00:01:58,416 –> 00:02:02,000
douwe_maan: does today, but especially at nine, as I’m sure you can imagine. Um. So
38
00:02:02,166 –> 00:02:06,416
douwe_maan: throughout my entire you know high school period from like eleven to eighteen, I
39
00:02:06,541 –> 00:02:10,625
douwe_maan: was doing freelance Web development. I figured out relatively quickly that this
40
00:02:10,791 –> 00:02:14,166
douwe_maan: skill that I was teaching myself Uh, was actually marketable and if you didn’t
41
00:02:14,333 –> 00:02:17,750
douwe_maan: tell anyone how young you were, they were actually willing to pretend that you
42
00:02:17,833 –> 00:02:22,166
douwe_maan: were an adult. They could go in business with Uh. Even, Um, when I was sixteen,
43
00:02:22,416 –> 00:02:26,791
douwe_maan: I ended up joining an IOS and Mac at the development studio at
44
00:02:27,041 –> 00:02:31,750
douwe_maan: Engineering leads, Um, and two years later with one of my founders or one of my
45
00:02:31,750 –> 00:02:36,416
douwe_maan: bosses. At the time I ended up founding a start up, Uh, when I was eighteen
46
00:02:36,958 –> 00:02:40,541
douwe_maan: where we were building software for bed and breakfasts to manage their
47
00:02:40,708 –> 00:02:44,458
douwe_maan: reservations, their guest communications, their website, et cetera, Um, And it
48
00:02:44,458 –> 00:02:48,875
douwe_maan: is through this company that I, ▁ultimately, ended up at Gitlab, Um. During my
49
00:02:49,041 –> 00:02:53,125
douwe_maan: time at this, this started I had. then I started studying computer science and
50
00:02:53,208 –> 00:02:56,458
douwe_maan: the Netherlands. Uh, I quickly realized that I didn’t necessarily need to give
51
00:02:56,541 –> 00:02:59,750
douwe_maan: that my full time attention because I’d already picked up so much of along the
52
00:02:59,833 –> 00:03:04,625
douwe_maan: way, so throughout my entire pes, um degree, I tried to go to as many
53
00:03:04,875 –> 00:03:08,708
douwe_maan: conferences around Ruby on Rails, the programming language that I was working
54
00:03:08,875 –> 00:03:14,791
douwe_maan: most with Eth time as I could. At some point I was in Ethens at a the European
55
00:03:15,125 –> 00:03:19,583
douwe_maan: Ruby conference and over lunch I ended up talking to a guy who was also standing
56
00:03:19,750 –> 00:03:23,291
douwe_maan: by himself at a table and introduced myself and I told him that I worked for
57
00:03:23,666 –> 00:03:28,083
douwe_maan: well the business I had, and that I was from the Netherlands. Um, and he ended
58
00:03:28,166 –> 00:03:30,958
douwe_maan: up pointing me towards his boss, who was in the other corner of the room,
59
00:03:31,208 –> 00:03:34,541
douwe_maan: telling him his boss were Dutch as well, and that we should go chat. So I
60
00:03:34,541 –> 00:03:37,750
douwe_maan: introduced myself to his boss and I told him about this company I had or rebuilt
61
00:03:37,916 –> 00:03:41,583
douwe_maan: Soware, Ben and Breakfasts, and as it turned out, this person I was talking to
62
00:03:41,750 –> 00:03:46,416
douwe_maan: was a s. C brandy. Now the C, e o and co founder of Getlab and his parents had a
63
00:03:46,416 –> 00:03:47,833
douwe_maan: bed and breakfast in the north of the
64
00:03:47,916 –> 00:03:52,000
douwe_maan: Netherlands, so just through this kind of chance, Uh meeting at a conference in
65
00:03:52,083 –> 00:03:56,333
douwe_maan: Athens, Um, he and I stated touch. He got an idea of what I was capable of
66
00:03:56,416 –> 00:03:59,833
douwe_maan: building even at a relatively young age, and we kept running into each other at
67
00:04:00,000 –> 00:04:03,833
douwe_maan: different meet ups and conferences around Europe and the netherences over the
68
00:04:03,833 –> 00:04:07,916
douwe_maan: coming months, to the extent where he asked me one day if I uh, wanted to come
69
00:04:08,083 –> 00:04:12,333
douwe_maan: join Get Lap, which at the time was a still mostly an open source project with a
70
00:04:12,416 –> 00:04:16,250
douwe_maan: community of several hundred open source contributors, but there was already a
71
00:04:16,333 –> 00:04:19,916
douwe_maan: team of about six in full time employees at the time trying to build a company
72
00:04:20,250 –> 00:04:24,333
douwe_maan: around. This, uh, this software for software development teams. At the time, I
73
00:04:24,458 –> 00:04:28,333
douwe_maan: was still very happy with my own start up and I, uh, I told Sid I wasn’t
74
00:04:28,458 –> 00:04:33,041
douwe_maan: interested six months later in January of Uh, twenty sixteen, not doing the
75
00:04:33,041 –> 00:04:35,500
douwe_maan: mathroon January of twenty fifteen, Um,
76
00:04:36,708 –> 00:04:40,083
douwe_maan: I reached out to sit again and I told him that Uh, my situation had changed and
77
00:04:40,083 –> 00:04:43,291
douwe_maan: he basically hired me on the spot And this was right when Getlap was going
78
00:04:43,416 –> 00:04:46,250
douwe_maan: through the White Combinator, a accelerator program in Mountain
79
00:04:46,458 –> 00:04:50,375
douwe_maan: View, California, Um, and gitlap was ten per people when I joined and about to
80
00:04:50,375 –> 00:04:51,083
douwe_maan: View, California, Um, and gitlap was ten per people when I joined and about to
81
00:04:51,291 –> 00:04:54,916
douwe_maan: ha, raise its first seed round of funding. So I joined GitLab as employee number
82
00:04:54,916 –> 00:04:55,541
douwe_maan: ha, raise its first seed round of funding. So I joined GitLab as employee number
83
00:04:55,791 –> 00:04:59,208
douwe_maan: ten. No idea where it could go And for those who haven’t seen the news,
84
00:04:59,291 –> 00:05:03,291
douwe_maan: Getitlab actually went public on the Naste Stock Exchange in October, So it’s
85
00:05:03,375 –> 00:05:07,166
douwe_maan: been an incredible journey that I had the privilege of seeing Uh from the front
86
00:05:07,375 –> 00:05:09,541
douwe_maan: lines from this very early stage. Um. During my time at Get Lab really quickly,
87
00:05:09,583 –> 00:05:11,291
douwe_maan: lines from this very early stage. Um. During my time at Get Lab really quickly,
88
00:05:11,458 –> 00:05:16,500
douwe_maan: I became the first engineering lead and manager at Getlab Uh, I ran the at the
89
00:05:16,500 –> 00:05:20,750
douwe_maan: time fourteen people engineering team for the first uh year or so. Um. And as
90
00:05:20,833 –> 00:05:24,333
douwe_maan: the company kept growing, I, I stayed in engineering leadership, and Um,
91
00:05:25,083 –> 00:05:29,000
douwe_maan: learning what it took for my team to be productive and learning also how the
92
00:05:29,083 –> 00:05:33,375
douwe_maan: tool that we were building get like Arself. was enabling teams to adopt these
93
00:05:33,458 –> 00:05:38,583
douwe_maan: devops === best practices in a way, Uh, that was never possible, or s as accessible
94
00:05:39,166 –> 00:05:40,583
douwe_maan: before. At some point though, when gitlap had reached fourteen hundred people, I
95
00:05:40,583 –> 00:05:42,500
douwe_maan: before. At some point though, when gitlap had reached fourteen hundred people, I
96
00:05:42,500 –> 00:05:46,208
douwe_maan: was starting to feel that itch of of wanting to work at a smaller company again,
97
00:05:46,583 –> 00:05:47,250
douwe_maan: a a more um, nascent project where you can have a really massive impact with
98
00:05:47,250 –> 00:05:49,875
douwe_maan: Um, Yeah, I, I wanted to join a smaller project again, Uh, and I thought that
99
00:05:49,958 –> 00:05:53,958
douwe_maan: that would require me leaving Get lab. Which I was hard pressed to do because of
100
00:05:54,125 –> 00:05:59,208
douwe_maan: getlabs kind of unique approach to all remote work, and a lot of flexibility
101
00:05:59,458 –> 00:06:00,833
douwe_maan: around how people combine their work
102
00:06:01,083 –> 00:06:04,916
douwe_maan: and their uh life and the freedom to travel, which is something that I can talk
103
00:06:05,083 –> 00:06:08,125
douwe_maan: about a little bit more, which has been really important. Uh in in my life so
104
00:06:08,166 –> 00:06:11,625
douwe_maan: far, Um, I live in Mexico City, for what it’s worth, not in the netherences
105
00:06:11,708 –> 00:06:15,291
douwe_maan: anymore. Uh, so there’s a story there, too, So I was at this point where I
106
00:06:15,375 –> 00:06:17,875
douwe_maan: thought I would need to leave Getlab to find what I was looking for, and this
107
00:06:17,958 –> 00:06:21,625
douwe_maan: kind of unique opportunity came up to join the Melano project, which was an
108
00:06:21,791 –> 00:06:25,958
douwe_maan: internal project in Getlab that was founded into Thousand Eighteen Um by the Get
109
00:06:26,125 –> 00:06:29,708
douwe_maan: Up Data team, who realized that a lot of these software development best
110
00:06:29,875 –> 00:06:34,250
douwe_maan: practices that they saw day and day out in the internal team with Ki Lab, as
111
00:06:34,250 –> 00:06:38,583
douwe_maan: well as at Getlap’s customers, could similarly benefit data teams. Uh, and make
112
00:06:38,750 –> 00:06:42,916
douwe_maan: them more collaborative, give them higher confidence to experiment and try stuff
113
00:06:43,125 –> 00:06:46,666
douwe_maan: out without fear of breaking things, and in the end increased their confidence
114
00:06:47,000 –> 00:06:51,708
douwe_maan: in the results of of the work that they do, and uh the data that their business
115
00:06:52,125 –> 00:06:56,166
douwe_maan: Uh, uses to make better decisions and and and improve their products. So I
116
00:06:56,166 –> 00:06:57,458
douwe_maan: joined Notano, Um,
117
00:06:58,500 –> 00:07:02,500
douwe_maan: kind of to bring that software engineering experience into the The data tool
118
00:07:02,750 –> 00:07:07,125
douwe_maan: space. Um. At the time, Montana had been around S for about a year and a half,
119
00:07:07,375 –> 00:07:11,000
douwe_maan: and even though we had this broad vision of what it could be, we had not really
120
00:07:11,125 –> 00:07:14,833
douwe_maan: been able to find detraction and the enthusiasm in the community that we were
121
00:07:14,916 –> 00:07:19,125
douwe_maan: looking for, so the decision was made at the time to um take the six person team
122
00:07:19,375 –> 00:07:23,625
douwe_maan: that existed and basically reduced the head countt on the project from six
123
00:07:23,791 –> 00:07:27,791
douwe_maan: people down to one. To give ourselves a little bit more runway, to figure out if
124
00:07:27,875 –> 00:07:31,125
douwe_maan: we could piv it and and find something that would really resonate with these
125
00:07:31,208 –> 00:07:35,875
douwe_maan: data, people that we were trying to uh to empower Um. So in early twenty twenty,
126
00:07:36,041 –> 00:07:39,541
douwe_maan: I was left on the Montana project by myself. Uh, and I had to figure out
127
00:07:39,625 –> 00:07:43,791
douwe_maan: relatively quickly how to turn it Ro. uh. Fortunately I did. We can talk about
128
00:07:43,875 –> 00:07:47,958
douwe_maan: that more at length, the in duration of the call, Um, to the extent that by the
129
00:07:47,958 –> 00:07:52,125
douwe_maan: end of twenty twenty, Gitlab had enough confidence in the project that I could
130
00:07:52,125 –> 00:07:56,041
douwe_maan: bring some more people onto the team, So I hired Um. one of the original people
131
00:07:56,125 –> 00:07:59,000
douwe_maan: that had been involved with the Moltana project. Internally, it get Lap, Tayor,
132
00:07:59,125 –> 00:08:03,625
douwe_maan: Murphy was our head of product and data, as well as one of the most active
133
00:08:03,875 –> 00:08:08,041
douwe_maan: contributors from the open source community. A ▁j steers, who was now our head
134
00:08:08,166 –> 00:08:12,041
douwe_maan: of engineering, and around this time I was talking to the Gillap C Eo, about
135
00:08:12,125 –> 00:08:17,083
douwe_maan: where G. Notana could go, and I, um, we realized we came to the conclusion that
136
00:08:17,125 –> 00:08:20,416
douwe_maan: in order for Moltana to really reach its full potential, spitting it out as an
137
00:08:20,500 –> 00:08:24,333
douwe_maan: independent Startu, raising some external funding and really being able to go
138
00:08:24,500 –> 00:08:28,666
douwe_maan: all in on this vision we had, without being in some sense limited in our
139
00:08:28,833 –> 00:08:32,583
douwe_maan: aspirations, because of the bigger machine that get up had become. I was the
140
00:08:32,583 –> 00:08:37,291
douwe_maan: best path forward, so in May we spun out of GitLab, Uh, which we announced in
141
00:08:37,375 –> 00:08:41,875
douwe_maan: June, and the team has now grown to ten people, and we have a lot of plans for
142
00:08:41,875 –> 00:08:42,875
douwe_maan: the rest of the year.
143
00:08:43,750 –> 00:08:48,000
anthony_algmin: thats. it’s such a cool story already, and I feel like we’re just at the
144
00:08:48,041 –> 00:08:52,875
anthony_algmin: very beginning of where this is going to hand. And there are so many things
145
00:08:53,083 –> 00:08:57,833
anthony_algmin: that idlicateated a candy store right now now, because this is the world I
146
00:08:58,000 –> 00:09:02,375
anthony_algmin: live in and I’ve seen in different perspectives. Um, and it doesn’t really
147
00:09:02,791 –> 00:09:08,875
anthony_algmin: surprise me that Meltao as part of of Git lab, and and with G it it’s all
148
00:09:09,750 –> 00:09:13,750
anthony_algmin: that is a different side of the house. Traditionally, which is the side of
149
00:09:13,750 –> 00:09:17,083
anthony_algmin: the house I come from? Is the data spacease right? And like we couldn’t get
150
00:09:17,250 –> 00:09:22,375
anthony_algmin: people to, do you know, code management and checkens, let alone a more
151
00:09:22,708 –> 00:09:27,500
anthony_algmin: comprehensive collaborative uh of way of working. So there’s some things
152
00:09:27,583 –> 00:09:33,750
anthony_algmin: that I want to dig into on that as well, but I can definitely see where it
153
00:09:34,000 –> 00:09:37,250
anthony_algmin: once. Given a little bit of distance from that application development side
154
00:09:37,416 –> 00:09:40,541
anthony_algmin: of the house, the code development side of the house, where if you can
155
00:09:40,708 –> 00:09:44,958
anthony_algmin: evolve what you’re offering in Meltono to really fit the data community,
156
00:09:46,041 –> 00:09:49,666
anthony_algmin: I could see it being tremendously valuable, and I’m seeing in in a number of
157
00:09:49,750 –> 00:09:54,708
anthony_algmin: different capacities. Uh, the the notion and and the practices around Data
158
00:09:54,875 –> 00:09:58,625
anthony_algmin: Os really starting to hit stride now, so I think from a market perspective,
159
00:09:59,083 –> 00:10:03,166
anthony_algmin: Uh, the the world is your oyster. You have so much potential here, so can
160
00:10:03,500 –> 00:10:08,875
anthony_algmin: for the people who are not as detailed and familiar with what even uh Gitlab
161
00:10:09,083 –> 00:10:15,833
anthony_algmin: does. Can you talk a little bit about what the origin of Meltao was? And and
162
00:10:15,916 –> 00:10:19,666
anthony_algmin: kind of where how its spun out of that? What? What get liberally does? What
163
00:10:19,833 –> 00:10:24,875
anthony_algmin: Meltao kind of was in the very beginning, and kind of where you see it going
164
00:10:25,166 –> 00:10:29,083
anthony_algmin: from here. And and what your kind of unique market, uh position is going to
165
00:10:29,083 –> 00:10:30,083
anthony_algmin: be with this company
166
00:10:30,166 –> 00:10:35,458
douwe_maan: Yeah, absolutely, and I agree with you that the the timing for day ups tooling
167
00:10:35,875 –> 00:10:39,791
douwe_maan: has has never been better. So we are a really nice infction point here and how
168
00:10:39,791 –> 00:10:44,125
douwe_maan: we think the data space will look, and the unique background in Get Lab that we
169
00:10:44,166 –> 00:10:49,625
douwe_maan: have, I think will bring significance, um, gap value to the table for us to
170
00:10:49,708 –> 00:10:53,708
douwe_maan: actually make this happen, So to provide that context that can be useful for the
171
00:10:53,708 –> 00:10:59,708
douwe_maan: audience, Get lab, Um is a an endtentth platform for the entire software
172
00:10:59,958 –> 00:11:03,875
douwe_maan: development life cycle, So if you have a company that builds software and this
173
00:11:03,958 –> 00:11:07,458
douwe_maan: can be a dedicated tech company, or just a the software division at any large
174
00:11:07,458 –> 00:11:08,458
douwe_maan: company. These
175
00:11:08,458 –> 00:11:11,958
douwe_maan: days that maintains some amount of internal toling you are talking about
176
00:11:12,125 –> 00:11:16,250
douwe_maan: software engineers. you probably have product managers. You have Um, user
177
00:11:16,416 –> 00:11:19,958
douwe_maan: interface designers and a whole bunch of different disciplines that come
178
00:11:20,166 –> 00:11:23,791
douwe_maan: together to build high quality software. And there’s a life cycle to the way
179
00:11:23,958 –> 00:11:27,083
douwe_maan: that this process works. Where you usually start with some kind of idea, or you
180
00:11:27,125 –> 00:11:30,916
douwe_maan: get user feedback on something you want to build. Then the next step is to spck
181
00:11:31,375 –> 00:11:34,583
douwe_maan: what this can look like in terms of additional functionality in the product that
182
00:11:34,583 –> 00:11:39,541
douwe_maan: you’re building. Then there is usually some stage where Um, other stakeholders
183
00:11:39,708 –> 00:11:43,625
douwe_maan: like these interfacees, designers come in to flash out this this idea, and then
184
00:11:43,791 –> 00:11:47,208
douwe_maan: there’s the development state where the actual software engineers gets started
185
00:11:47,541 –> 00:11:51,083
douwe_maan: and software engineers don’t just start writing code and then finish by the end
186
00:11:51,125 –> 00:11:54,500
douwe_maan: of the day, and everything’s perfect. It is a usually iterative process
187
00:11:54,833 –> 00:11:58,250
douwe_maan: where it’s starting from fundamentals, making many, many, many changes, and
188
00:11:58,333 –> 00:12:03,208
douwe_maan: trying to get incrementally closer to this this goal and in order to have a lot
189
00:12:03,375 –> 00:12:07,291
douwe_maan: of confidence in what you’re building. Uh, both it meeting the requirements and
190
00:12:07,291 –> 00:12:10,250
douwe_maan: you’re not accidentally breaking things. Um, you
191
00:12:10,666 –> 00:12:15,000
douwe_maan: need tooling that helps with change management. that helps collaborate in teams
192
00:12:15,083 –> 00:12:20,041
douwe_maan: of multiple engineers. It helps with automatic validation of Um, the Code of
193
00:12:20,041 –> 00:12:21,041
douwe_maan: Britain, and, and making
194
00:12:21,125 –> 00:12:24,166
douwe_maan: sure that functionality rorote last year doesn’t accidentally break Because you
195
00:12:24,250 –> 00:12:28,041
douwe_maan: made a change somewhere that has some in in unattended impact. And then there
196
00:12:28,125 –> 00:12:32,125
douwe_maan: are further stages where through the process of writing this code and getting it
197
00:12:32,166 –> 00:12:36,666
douwe_maan: to that level where it passes some checks, you can automatically deploy it onto
198
00:12:37,125 –> 00:12:41,458
douwe_maan: Um, users devices, or the the The website where users will get to interact with
199
00:12:41,541 –> 00:12:44,416
douwe_maan: it, And this is a life cycle because then you usually go back to the first step
200
00:12:44,583 –> 00:12:47,708
douwe_maan: again. You start working on your next idea or an improvement to the original
201
00:12:48,125 –> 00:12:52,041
douwe_maan: idea, and Glab is a platform that brings all of this functionality that Soware
202
00:12:52,166 –> 00:12:56,333
douwe_maan: development teams need to collborate into one massive platform that then becomes
203
00:12:56,500 –> 00:13:01,291
douwe_maan: kind of like an operating system or the one tool a software Hadp team lives in
204
00:13:01,541 –> 00:13:06,750
douwe_maan: Um. That enables them to build really high quality of software together, Um, and
205
00:13:07,000 –> 00:13:11,375
douwe_maan: do so more quickly than Uh. Without you know the help the platform provides. So
206
00:13:11,791 –> 00:13:16,166
douwe_maan: Gitlab came from an open source project called Gitlab, built around an open
207
00:13:16,333 –> 00:13:20,250
douwe_maan: source technological gift, which is also the open source ofology, that Getis Hub
208
00:13:20,416 –> 00:13:23,208
douwe_maan: is built around which you might be familiar with. They were acquired by
209
00:13:23,208 –> 00:13:27,000
douwe_maan: Microsoft a number of years ago and they do something very similar, but there is
210
00:13:27,125 –> 00:13:32,416
douwe_maan: a significant difference in that Github, itself, Um, was a proprietary product
211
00:13:32,583 –> 00:13:36,125
douwe_maan: that you buy and you pay for and new use and you can use it within your own
212
00:13:36,250 –> 00:13:40,250
douwe_maan: organization. While Gitlab from the one has been open source, so all of the
213
00:13:40,333 –> 00:13:43,541
douwe_maan: source code has always been available for free to the whole world, And it was
214
00:13:43,708 –> 00:13:48,333
douwe_maan: very much a collaborate effort, collaborative effort of a really large software
215
00:13:48,500 –> 00:13:52,833
douwe_maan: development community around the entire world that came together to work on this
216
00:13:53,000 –> 00:13:56,250
douwe_maan: tool together because they wanted it themselves. they knew it would benefit
217
00:13:56,750 –> 00:13:59,958
douwe_maan: others, and they saw an opportunity to bring all of their own ideas and
218
00:14:00,041 –> 00:14:04,666
douwe_maan: perspectives and and mistakes of the past, Uh together to build their ideal
219
00:14:04,916 –> 00:14:09,083
douwe_maan: platform to make themselves more productive, Um. And very much, building on this
220
00:14:09,125 –> 00:14:14,250
douwe_maan: philosophy of building tools together with the users of those tools, and the
221
00:14:14,416 –> 00:14:19,375
douwe_maan: philosophy of Um, enabling more effective collaboration by having automated
222
00:14:19,541 –> 00:14:24,125
douwe_maan: tests, by having some amount of of change management and code reviews that
223
00:14:24,166 –> 00:14:27,291
douwe_maan: people can make changes with confidence that they won’t break anything, Because
224
00:14:27,458 –> 00:14:30,750
douwe_maan: someone else in the team can come in and bring in feedback. The data team at
225
00:14:30,833 –> 00:14:35,125
douwe_maan: Moano, rather, the data team at Kidlap, realize that these were qualities that
226
00:14:35,208 –> 00:14:40,125
douwe_maan: data teams would similarly benefit from, and Um. They decided to start building
227
00:14:40,250 –> 00:14:44,583
douwe_maan: open source data tools that would try to encompass the entire data life cycle
228
00:14:45,125 –> 00:14:51,083
douwe_maan: and Um, do so while bringing in some of these Deth Os development operations
229
00:14:51,375 –> 00:14:55,541
douwe_maan: best practices that I’ve talked about a second ago like Version Control Code
230
00:14:55,625 –> 00:14:59,708
douwe_maan: review and continuous integration and employment, which is fancy words. that
231
00:14:59,791 –> 00:15:03,958
douwe_maan: basically mean automated testing and nothing goes life until all the check boxes
232
00:15:04,125 –> 00:15:08,500
douwe_maan: are checked, Um, and wrapping this all up in in in tolling, that would
233
00:15:08,750 –> 00:15:12,166
douwe_maan: revolutionize the data spacease in the same sense, Deaf Ose software
234
00:15:12,166 –> 00:15:13,166
douwe_maan: development.
235
00:15:15,125 –> 00:15:16,125
anthony_algmin: that’s awesome.
236
00:15:17,125 –> 00:15:18,125
anthony_algmin: Just so,
237
00:15:18,958 –> 00:15:24,625
anthony_algmin: can you talk a little bit about the differences and the way data people work
238
00:15:25,000 –> 00:15:30,458
anthony_algmin: versus the software engineering side of the house, And like what you need to
239
00:15:30,541 –> 00:15:34,458
anthony_algmin: address differently is melt versus what Getitlap would do for first
240
00:15:34,500 –> 00:15:35,500
anthony_algmin: softwareers
241
00:15:36,041 –> 00:15:40,333
douwe_maan: Yeah, great question. uh, there are a number of differences that are relevant,
242
00:15:40,500 –> 00:15:44,416
douwe_maan: but the thing that surprised me the most is how similar the state of the data
243
00:15:44,666 –> 00:15:47,791
douwe_maan: spacease today is with the tooling available to it and the way teams collaborate
244
00:15:47,958 –> 00:15:52,250
douwe_maan: to where the software development space was maybe ten years ago. And looking
245
00:15:52,416 –> 00:15:56,666
douwe_maan: back at what Gitlap is doing, we realized that very early on, the biggest
246
00:15:56,833 –> 00:16:00,833
douwe_maan: challenge wasn’t necessarily in convincing companies to use Getlab for their
247
00:16:01,083 –> 00:16:04,500
douwe_maan: software development life cycle. The problem was convincing them that they
248
00:16:04,666 –> 00:16:08,250
douwe_maan: wanted develops at all that they wanted automated testing in automated
249
00:16:08,416 –> 00:16:12,333
douwe_maan: deployment at all, and that the waterfall based way of working in the past,
250
00:16:12,416 –> 00:16:16,250
douwe_maan: where it’s very much about different disciplines doing their work to perfection
251
00:16:16,416 –> 00:16:21,208
douwe_maan: and handing it over to the next team. Um didn’t work, and with the agile
252
00:16:21,375 –> 00:16:25,875
douwe_maan: methodogies of collaboration and and project management finding ground in
253
00:16:25,958 –> 00:16:29,875
douwe_maan: software development, the types of tooling that these teams needed also changed.
254
00:16:30,125 –> 00:16:33,000
douwe_maan: but in the beginning it getlap. we still to convince people they wanted the
255
00:16:33,083 –> 00:16:36,916
douwe_maan: stuff at all, and to these days that is kind of a given. and we see something
256
00:16:37,125 –> 00:16:40,583
douwe_maan: similar happening with data ups Where the challenge today isn’t even so much
257
00:16:40,666 –> 00:16:43,875
douwe_maan: about convincing people they want Montana, but convincing them that they won Da
258
00:16:43,958 –> 00:16:49,000
douwe_maan: Dos at all. It really means, which really means convincing data teams that they
259
00:16:49,083 –> 00:16:52,166
douwe_maan: are far more similar to software development teames then they might even
260
00:16:52,333 –> 00:16:56,041
douwe_maan: realize, and that they would actually benefit from Steing the work day. Do and
261
00:16:56,250 –> 00:16:59,708
douwe_maan: uh, you know the the data stack that they build by picking different tools and
262
00:16:59,791 –> 00:17:03,458
douwe_maan: lookoking them together and building custom components, And then you know having
263
00:17:03,625 –> 00:17:07,958
douwe_maan: a dashboard or some kind of notebook that a data scientist or Da liist works in,
264
00:17:08,333 –> 00:17:11,958
douwe_maan: We want them to see that this is way more similar to product building than they
265
00:17:11,958 –> 00:17:15,541
douwe_maan: might realize, And their data Stck is not just a combination of of purchasing
266
00:17:15,708 –> 00:17:18,500
douwe_maan: decisions, Like I’ll choose these five products and that’s to will use forever.
267
00:17:18,833 –> 00:17:24,041
douwe_maan: It is really you, building a bespoke piece of software of a spoke product suited
268
00:17:24,166 –> 00:17:28,583
douwe_maan: to the needs of the specific organization where your users are, your, your, your
269
00:17:28,666 –> 00:17:32,583
douwe_maan: colleagues in the company that wanna somehow benefit from the data that you make
270
00:17:32,625 –> 00:17:33,625
douwe_maan: available to them.
271
00:17:33,625 –> 00:17:36,916
douwe_maan: The features of the product are any way that they have of interacting with it.
272
00:17:37,083 –> 00:17:41,291
douwe_maan: So those are visualization methods, Uh, the U, the B, I, and Ellythtic tools and
273
00:17:41,541 –> 00:17:46,125
douwe_maan: Um. opportunities to run. You know, Jupiter, Notebooks and the data team, The
274
00:17:46,125 –> 00:17:50,583
douwe_maan: data engineers and analytics engineers. Their task is really to build a product
275
00:17:50,750 –> 00:17:55,875
douwe_maan: and continuously improve it day after day to more successfully meet Um, the
276
00:17:55,958 –> 00:17:58,583
douwe_maan: needs of their colleagues and the goals of the organization, and all of these
277
00:17:58,666 –> 00:18:02,125
douwe_maan: advantages of software development that have talked about where you wa to have
278
00:18:02,250 –> 00:18:05,000
douwe_maan: really high confidence that when you want to bring make a change, you don’t
279
00:18:05,125 –> 00:18:08,500
douwe_maan: accidentally break something down the line. A software development and defils
280
00:18:08,666 –> 00:18:12,250
douwe_maan: have answers to that that the data spacease and data tooling today isn’t really
281
00:18:12,333 –> 00:18:16,416
douwe_maan: set up for, because many times you end up with some Sas software running in your
282
00:18:16,500 –> 00:18:20,416
douwe_maan: browser and you make a change there. and the I. the effect immediately, Um
283
00:18:20,833 –> 00:18:24,041
douwe_maan: affects. You know. your c, f O was about to do a board meeting and wants to
284
00:18:24,125 –> 00:18:27,291
douwe_maan: present their dashboard, And if you accidentally made typo, then the live
285
00:18:27,541 –> 00:18:31,875
douwe_maan: Dashbard will be broken. So coming from the software development world, for, I
286
00:18:31,958 –> 00:18:35,375
douwe_maan: really haven’t been much exposed to data teams and the way they work until I
287
00:18:35,541 –> 00:18:38,833
douwe_maan: started looking into Milanno, and then I had this opportunity to join the team.
288
00:18:39,083 –> 00:18:44,250
douwe_maan: I was really surprised to see that the parallels to me were obvious. Data teams
289
00:18:44,500 –> 00:18:48,125
douwe_maan: themselves might not even realize that those barriels existed, but for me this
290
00:18:48,166 –> 00:18:52,041
douwe_maan: was just an opportunity to bring all of these ten years worth of lessons from
291
00:18:52,125 –> 00:18:55,791
douwe_maan: software development and the specific lessons we had learned at get building
292
00:18:55,958 –> 00:19:02,250
douwe_maan: this open source developer tools, Um. For and by software engineers into the
293
00:19:02,333 –> 00:19:07,458
douwe_maan: data spacee, Um, some of the differences, though one big difference is that over
294
00:19:07,625 –> 00:19:12,416
douwe_maan: the last five to ten years the data spacee has become incredibly horizontally
295
00:19:12,583 –> 00:19:16,833
douwe_maan: integrated when it comes to the tooling, so gone are the days of these big
296
00:19:17,083 –> 00:19:20,583
douwe_maan: integrated intertuent data platforms that companies were using them fiftyft
297
00:19:20,916 –> 00:19:23,958
douwe_maan: years ago, and the day companies are still using today. Instead, you have the
298
00:19:24,041 –> 00:19:27,291
douwe_maan: modern data stack where you make like six or seven tooling decisions to try to
299
00:19:27,375 –> 00:19:30,333
douwe_maan: get the best word, the task at hand, And then there iss a challenge for a
300
00:19:30,333 –> 00:19:34,416
douwe_maan: specific data team in hooking these up together and and making it all work
301
00:19:34,750 –> 00:19:35,875
douwe_maan: smoothly. Um,
302
00:19:36,833 –> 00:19:41,291
douwe_maan: what that means is that this vision we had of building an intuent platform for
303
00:19:41,291 –> 00:19:45,083
douwe_maan: the data life cycle cannot be realized just by building one massive tool and
304
00:19:45,125 –> 00:19:49,125
douwe_maan: saying, replace everything you currently have with this new thing Because as a
305
00:19:49,208 –> 00:19:53,125
douwe_maan: consequence of this versontal integration at every step of the life cycle at
306
00:19:53,208 –> 00:19:57,625
douwe_maan: every layer of the data stack, there are so many kind of really great tools
307
00:19:57,875 –> 00:20:01,791
douwe_maan: competing, and and focused so narrowly on a really specific problem to be
308
00:20:01,875 –> 00:20:06,833
douwe_maan: solved, that a single product can never really claim to do it all better or
309
00:20:07,000 –> 00:20:10,500
douwe_maan: even. Good enough. compared with the tools that spend every day just a training
310
00:20:10,666 –> 00:20:14,166
douwe_maan: on a particular area with Giitlab. the situation was a little bit different ten
311
00:20:14,333 –> 00:20:19,291
douwe_maan: years ago. Um, because there wasn’t really any platform that try to.
312
00:20:20,583 –> 00:20:23,458
douwe_maan: well, First of all, there wasn’t the platform that tried to do everything for
313
00:20:23,458 –> 00:20:27,458
douwe_maan: the entire dietal life cycle, But the horizontal integration also hadn’t evolved
314
00:20:27,541 –> 00:20:30,583
douwe_maan: to the extent that you have multiple really great competing solutions in every
315
00:20:30,750 –> 00:20:34,500
douwe_maan: step. So Gitlap’s approach was to just build one entertent platform and all of
316
00:20:34,500 –> 00:20:38,416
douwe_maan: the contributors in an open source ecosystem to contribute to this one tool. and
317
00:20:38,416 –> 00:20:41,791
douwe_maan: then we will build something amazing and get the best been able to do that in
318
00:20:42,041 –> 00:20:44,666
douwe_maan: data, because things are a little bit more fragmented and horizontally
319
00:20:44,750 –> 00:20:48,583
douwe_maan: integrated. Our opportunity is different. Uh. We have realized that in order to
320
00:20:48,625 –> 00:20:49,625
douwe_maan: bring develop best
321
00:20:49,791 –> 00:20:53,791
douwe_maan: practices to data which you can summarize as data, Os, uh, what’s really needed
322
00:20:53,958 –> 00:20:58,041
douwe_maan: is a new layer in the data stack, a foundational layer. If you will, would like
323
00:20:58,125 –> 00:21:01,625
douwe_maan: to call ourselves the foundation of Every’s ideal datatack, Because we
324
00:21:01,708 –> 00:21:05,125
douwe_maan: acknowledge that every’ data Se, a complete difference is going to look
325
00:21:05,291 –> 00:21:08,666
douwe_maan: different today versus six months from now and five years from now. All of the
326
00:21:08,750 –> 00:21:12,583
douwe_maan: tools might have been swapted out, but there is still advantage in having some
327
00:21:13,208 –> 00:21:17,375
douwe_maan: um part of that stack that brings it all together and allows teams to interact
328
00:21:17,541 –> 00:21:23,625
douwe_maan: with it as if they are building one data product that they can easily you know,
329
00:21:23,875 –> 00:21:27,291
douwe_maan: run and test locally without accidentally affecting production. They can have
330
00:21:27,375 –> 00:21:32,125
douwe_maan: automated tests running somewhere in their Devil platform, so that they will see
331
00:21:32,166 –> 00:21:35,083
douwe_maan: an error message that accidentally broke something before actually getting into
332
00:21:35,375 –> 00:21:39,125
douwe_maan: production. But all of this requires this, this loose combination of tools to
333
00:21:39,375 –> 00:21:43,791
douwe_maan: still have some foundational area that that brings it together and allows it to
334
00:21:43,791 –> 00:21:47,458
douwe_maan: be treated as one unit, as that’s a massive difference in the approach that we
335
00:21:47,541 –> 00:21:53,000
douwe_maan: are taking with Notano as a modular open source data platform versus a Getitlap
336
00:21:53,125 –> 00:21:56,416
douwe_maan: which was an all in one open source De platform.
337
00:21:58,333 –> 00:21:59,333
anthony_algmin: So
338
00:22:00,625 –> 00:22:05,083
anthony_algmin: oh man, there’s so much that I could get out of my soap box about in what
339
00:22:05,166 –> 00:22:10,625
anthony_algmin: you about and I, I will just say, I will just say. I absolutely agree that
340
00:22:10,958 –> 00:22:14,291
anthony_algmin: it. It seems like it’s history repeating itself, but in the datase versus
341
00:22:14,625 –> 00:22:18,375
anthony_algmin: the software engineering space, and I would argue that even for the last
342
00:22:18,875 –> 00:22:23,583
anthony_algmin: several decades, the walls that have differentiated the data spacease from
343
00:22:23,666 –> 00:22:27,583
anthony_algmin: the software engineering space have been crumbling slowly but surely, and
344
00:22:27,750 –> 00:22:32,791
anthony_algmin: that there are things that have kind of thrown the line over that wall. I
345
00:22:32,875 –> 00:22:36,875
anthony_algmin: think of things like No sequel, which is now the domain of more data people
346
00:22:37,000 –> 00:22:40,291
anthony_algmin: than it used to be. where the data people were like. Well, we relational.
347
00:22:40,458 –> 00:22:44,125
anthony_algmin: this no sequel business, The application, folks that are literally had
348
00:22:44,291 –> 00:22:48,708
anthony_algmin: created it to do the kinds of things that the Web needed to be done. The
349
00:22:48,791 –> 00:22:53,333
anthony_algmin: search and retrieval of you. a lot of contextual information related to a
350
00:22:53,416 –> 00:22:58,708
anthony_algmin: specific account or specific person, specific entity that you know. it helps
351
00:22:59,000 –> 00:23:03,250
anthony_algmin: drive their engagement with that uh application, Uh, the web application or
352
00:23:03,333 –> 00:23:06,791
anthony_algmin: internet site. Um, you know it was. It was born out of necessity and
353
00:23:06,958 –> 00:23:10,458
anthony_algmin: eventually people on the data side are like. You know what. That’s pretty
354
00:23:10,708 –> 00:23:14,375
anthony_algmin: good at that particular thing. Like, Let’s understand what that happens and
355
00:23:14,458 –> 00:23:17,333
anthony_algmin: more data. People have gotten exposure to no sequal and have started to
356
00:23:17,333 –> 00:23:20,708
anthony_algmin: expand the horizons if you see similar patterns taking place with knowledge,
357
00:23:20,875 –> 00:23:25,250
anthony_algmin: graphs, and and other things that are the domain. Traditionally, I would
358
00:23:25,250 –> 00:23:28,875
anthony_algmin: argue of the application and software development side of the house, but I
359
00:23:29,000 –> 00:23:34,000
anthony_algmin: certainly agree with your assertions that the data space, if we cast a
360
00:23:34,000 –> 00:23:37,333
anthony_algmin: slightly different view than maybe the the data people traditionally would,
361
00:23:37,666 –> 00:23:40,208
anthony_algmin: especially enterprise data. People would have traditionally thought of
362
00:23:40,375 –> 00:23:44,875
anthony_algmin: themselves as it really is a software engineering application development
363
00:23:45,000 –> 00:23:48,791
anthony_algmin: type of function with just a few differences in what they’re working with.
364
00:23:48,958 –> 00:23:53,250
anthony_algmin: The their clay is different, but there’re still sculptors, and so I do find
365
00:23:53,416 –> 00:23:59,916
anthony_algmin: it. It’s very comparable and interesting to think about this as a. You know,
366
00:24:00,125 –> 00:24:02,708
anthony_algmin: it’s in some ways it it. It’s bringing together
367
00:24:03,833 –> 00:24:08,000
anthony_algmin: what are two sides of the coin already, and and just using more similar
368
00:24:08,291 –> 00:24:12,625
anthony_algmin: tools. Hey, we can learn from the software engineers who have solved this
369
00:24:12,791 –> 00:24:17,250
anthony_algmin: particular set of challenges before were not good at process and operations
370
00:24:17,416 –> 00:24:20,875
anthony_algmin: in the Dat spacease, where where we think about things andway it’s like we
371
00:24:21,000 –> 00:24:25,500
anthony_algmin: write ▁queries, using the same kind of iterative process that software
372
00:24:25,666 –> 00:24:29,833
anthony_algmin: engineers use to write code. But we don’t acknowledge the fact that there’s
373
00:24:30,000 –> 00:24:34,000
anthony_algmin: value in understanding those iterations. We try to make it so that that
374
00:24:34,125 –> 00:24:38,875
anthony_algmin: ▁querier those analytics get to the right answer, but then oftenims and this
375
00:24:38,875 –> 00:24:42,625
anthony_algmin: is where I do think there’s a difference oftenimes we. We don’t see the
376
00:24:42,791 –> 00:24:48,291
anthony_algmin: result from that until much later when a c f O is looking at two reports
377
00:24:48,458 –> 00:24:51,666
anthony_algmin: where the number should be the same and they’re different. Like, why is that
378
00:24:51,833 –> 00:24:55,000
anthony_algmin: and then you realize, Oh well, it, it was calculated differently, or the
379
00:24:55,166 –> 00:24:59,916
anthony_algmin: differences, s, or whatever it was. But there’s often that time delay
380
00:25:00,041 –> 00:25:03,666
anthony_algmin: because data starts to hide underneath the surface, which actually makes it
381
00:25:03,833 –> 00:25:08,708
anthony_algmin: that much more important to have rigor in the process of developing it.
382
00:25:09,000 –> 00:25:13,083
anthony_algmin: Because you want to catch these things up front. otherwise they remain
383
00:25:13,416 –> 00:25:16,791
anthony_algmin: buried. You’re not going to get that immediate feedback. Hey, we’re we’re
384
00:25:16,875 –> 00:25:21,333
anthony_algmin: going to do this presentation and Oh, the thing just broke. You’re going to
385
00:25:21,500 –> 00:25:25,916
anthony_algmin: find out later after you’ve already reported financial results to the market
386
00:25:26,208 –> 00:25:29,416
anthony_algmin: that now you’re going to have to go and restate things. Well, what’s more
387
00:25:29,666 –> 00:25:30,666
anthony_algmin: problematic?
388
00:25:31,500 –> 00:25:35,666
anthony_algmin: That kind of big problem that goes unseen for a long time or something, that
389
00:25:35,750 –> 00:25:39,666
anthony_algmin: ruins your morning with a bad demo, but can immediately be fixed and then
390
00:25:39,750 –> 00:25:42,208
anthony_algmin: you can move forward, And that’s what we want to do Is we want to reduce
391
00:25:42,375 –> 00:25:45,000
anthony_algmin: that cycle time. We want to get it to the point where we know immediately.
392
00:25:45,166 –> 00:25:49,250
anthony_algmin: Hey, we have a problem here. Let’s fix it versus having it be buried and
393
00:25:49,333 –> 00:25:51,166
anthony_algmin: then come up at the worst possible time. Later
394
00:25:51,875 –> 00:25:54,833
douwe_maan: Yeah, absolutely, and I would say that, Um,
395
00:25:56,041 –> 00:25:57,541
douwe_maan: things like having
396
00:25:58,750 –> 00:26:02,583
douwe_maan: consistent definitions of metrics so that you don’t have to. You know
397
00:26:02,666 –> 00:26:06,416
douwe_maan: contradictory numbers in the final dashboard and things like lineage, being able
398
00:26:06,500 –> 00:26:11,541
douwe_maan: to trace all the way back where some data point came from are um, go hand in
399
00:26:11,625 –> 00:26:15,875
douwe_maan: hand with this kind of platform way of thinking about data products like the
400
00:26:15,958 –> 00:26:21,458
douwe_maan: best data lineage tools today. Unfortunately, are these all in one platform? Is
401
00:26:21,541 –> 00:26:24,500
douwe_maan: that really do have everything under one roof, because they can easily relate
402
00:26:24,666 –> 00:26:28,333
douwe_maan: one day to point they to there. But if you are putting together a modern datatap
403
00:26:28,416 –> 00:26:31,791
douwe_maan: with five toling decisions, it becomes really difficult to have that information
404
00:26:32,166 –> 00:26:36,500
douwe_maan: be easily traceable between through the entire life cycle, so Motana, being kind
405
00:26:36,500 –> 00:26:40,416
douwe_maan: of an extrafational layer that has full visibility in entire datatack and all of
406
00:26:40,416 –> 00:26:45,083
douwe_maan: the different components also allows us to realize functionity like this that
407
00:26:45,208 –> 00:26:50,583
douwe_maan: previously would be hard to to realise in combination with the fact that since
408
00:26:50,833 –> 00:26:55,791
douwe_maan: we want to get data into the habit of versioning their change just and being
409
00:26:55,958 –> 00:26:59,291
douwe_maan: able to you know, describe with a Quickmit message why you made a certain
410
00:26:59,541 –> 00:27:03,291
douwe_maan: change, it will also be much more easy to trace back a particular error that
411
00:27:03,291 –> 00:27:07,208
douwe_maan: arose to the specific change that triggered it, which you can then trace back to
412
00:27:07,208 –> 00:27:10,125
douwe_maan: the Cod review conversation, where it was you know, signed off by another team
413
00:27:10,250 –> 00:27:13,458
douwe_maan: member, and they. Can trace expected specific issue that that prompted that
414
00:27:13,541 –> 00:27:17,708
douwe_maan: change to begin with, and just as this lineage it doesn’t just go through the Li
415
00:27:17,958 –> 00:27:21,625
douwe_maan: data through the data pipeline all the way to the point of data ingestion, but
416
00:27:21,708 –> 00:27:25,458
douwe_maan: actually also goes into the the meta level of the team. That that made this
417
00:27:25,625 –> 00:27:32,166
douwe_maan: change, and why is extremely powerful and seeing Um, your your data pipeline as
418
00:27:32,250 –> 00:27:36,583
douwe_maan: a type of software, Um, all the way from the ingestion piece to the the
419
00:27:36,583 –> 00:27:41,000
douwe_maan: Dashboard report definition, and the sequelqueries in Robia application, Um, is’
420
00:27:41,125 –> 00:27:44,250
douwe_maan: something that the software development space ten years ago hadn’t really caught
421
00:27:44,333 –> 00:27:47,291
douwe_maan: on too yet either, Like I remember when I started Know programming almost
422
00:27:47,375 –> 00:27:51,458
douwe_maan: eighteen years ago. now that I would just sign in over f t, p. To some life
423
00:27:51,625 –> 00:27:56,416
douwe_maan: production web serveurr download a ph P file, Make changes uploaded back and
424
00:27:56,500 –> 00:27:59,875
douwe_maan: then go through the main website to see if I did break something. and now it
425
00:28:00,333 –> 00:28:04,666
douwe_maan: unthinkable, but to me the way a lot of things work in Data Land still look like
426
00:28:04,916 –> 00:28:09,458
douwe_maan: something that I thought of as something of the past. But what I want to say?
427
00:28:09,625 –> 00:28:12,666
douwe_maan: what I wa to add to, kind of ask some coolor to what you were saying earlier
428
00:28:12,833 –> 00:28:16,250
douwe_maan: about Wow, Data can learn from software development. I think in the other
429
00:28:16,500 –> 00:28:20,666
douwe_maan: direction too. I think there’s a big opportunity for software engineering teams
430
00:28:20,833 –> 00:28:25,291
douwe_maan: and data teams to start feeling more like you said. Like kind of two sites of
431
00:28:25,291 –> 00:28:29,458
douwe_maan: the same coin, or or way more similar than they are different, Um, data
432
00:28:29,708 –> 00:28:33,083
douwe_maan: engineering and and data work more as a subset of software engineering and
433
00:28:33,083 –> 00:28:36,666
douwe_maan: anything else, Because what I’m also seeing is that software engineers that need
434
00:28:36,833 –> 00:28:41,708
douwe_maan: to do things like pulling data from various Ap, is transforming it in some way,
435
00:28:41,875 –> 00:28:46,125
douwe_maan: and in presenting it to the user in U. I. They don’t know what the right terms
436
00:28:46,250 –> 00:28:50,666
douwe_maan: are to even Google to find the E t, ▁l solutions that data teams have been
437
00:28:50,750 –> 00:28:55,208
douwe_maan: working with for decades, and we are trying to build software that can tool that
438
00:28:55,291 –> 00:28:58,750
douwe_maan: can really narry these worlds that will feel at home. The software developers
439
00:28:58,916 –> 00:29:02,583
douwe_maan: who already know that the only kind of Datap plan they would ever build is one
440
00:29:02,666 –> 00:29:06,583
douwe_maan: that’s going to be viual, controllable and festible, and at the same time
441
00:29:06,750 –> 00:29:10,750
douwe_maan: bringing these data engineers into these expectations by showing how how
442
00:29:10,916 –> 00:29:14,500
douwe_maan: beneficial has been for software, and then a place where we end up, hopefully is
443
00:29:14,666 –> 00:29:18,416
douwe_maan: where every software engineer will be able to be more producactive with data
444
00:29:18,583 –> 00:29:24,916
douwe_maan: work, and Everyda engineer will also feel more empower to Um, use software to to
445
00:29:25,125 –> 00:29:29,708
douwe_maan: you know, automate things beyond just writing data pipelines, because, just a
446
00:29:29,791 –> 00:29:33,708
douwe_maan: mental aspect. Of, I’m software engineer, your data person creates some
447
00:29:34,041 –> 00:29:38,125
douwe_maan: expectation in those people’s heads of what I can and can do, and and where the
448
00:29:38,125 –> 00:29:42,916
douwe_maan: limits of my skills and I think that’s completely unnecessary. And um, there’s
449
00:29:43,125 –> 00:29:47,958
douwe_maan: so much to be gained from these worlds kind colliding and you combining.
450
00:29:49,583 –> 00:29:53,500
anthony_algmin: couldn’t agree more, and and time spent trying to understand at a deeper
451
00:29:53,666 –> 00:29:56,000
anthony_algmin: level, what the other side of that coin
452
00:29:56,208 –> 00:30:00,000
anthony_algmin: does is really important. I’ll never forget. early in my consulting career,
453
00:30:00,291 –> 00:30:04,000
anthony_algmin: I was more on the data side. I came from a business intelligence and and you
454
00:30:04,000 –> 00:30:07,000
anthony_algmin: know, data warehousing world where we’re doing data modeling, and like,
455
00:30:07,250 –> 00:30:11,250
anthony_algmin: there’s a lot of complexity in creating these massive calculations and
456
00:30:11,416 –> 00:30:14,875
anthony_algmin: models to do things like financial reporting and things like that. There’ a
457
00:30:14,958 –> 00:30:18,041
anthony_algmin: lot of complexity there. And and I was really familiar with that and I
458
00:30:18,041 –> 00:30:21,000
anthony_algmin: didn’t note the software engineering side of the house as well, but I was
459
00:30:21,000 –> 00:30:24,208
anthony_algmin: working in a consulting firm where we had our our Aptdev team who were
460
00:30:24,375 –> 00:30:27,583
anthony_algmin: talking about writing to Ap I, which I knew a little bit about. But I hadn’t
461
00:30:27,666 –> 00:30:32,041
anthony_algmin: done a ton of in the data spacease at that time, and and I remember them
462
00:30:32,291 –> 00:30:36,458
anthony_algmin: just completely dismissing how complicated the data warehousing component
463
00:30:36,625 –> 00:30:38,875
anthony_algmin: was were like Well, we just need to pull out the data, but the data part’s
464
00:30:38,958 –> 00:30:41,750
anthony_algmin: easy, so we don’t have to worry about that and I’m like wait, wait. it is
465
00:30:41,916 –> 00:30:45,250
anthony_algmin: not easy. It is not something that you can just dismiss, But that’s human
466
00:30:45,416 –> 00:30:50,041
anthony_algmin: nature. We tend to devalue or diminish the complexity of the things that we
467
00:30:50,125 –> 00:30:54,708
anthony_algmin: don’t understand because we’re humans like. That’s just a fundamental trait
468
00:30:55,000 –> 00:30:59,750
anthony_algmin: that we have, but I do feel like when it comes to being a data, uh, person
469
00:31:00,125 –> 00:31:04,000
anthony_algmin: or a software engineering person and finding ways to realize that Hey, these
470
00:31:04,125 –> 00:31:09,166
anthony_algmin: are all interconnected. These are all just different facets of the same
471
00:31:09,666 –> 00:31:13,916
anthony_algmin: collection of challenges that we face to using data to improve our
472
00:31:14,000 –> 00:31:18,125
anthony_algmin: businesses. That’s really what data leadership is all about, and so having
473
00:31:18,958 –> 00:31:22,625
anthony_algmin: that awareness is a great first step. So, for anybody who’s listening and
474
00:31:22,708 –> 00:31:25,083
anthony_algmin: these concepts and the things that we’re talking about today with the with
475
00:31:25,166 –> 00:31:28,958
anthony_algmin: Devbs and Data Os, like you’ve got years worth of learning in front of you
476
00:31:29,083 –> 00:31:32,708
anthony_algmin: because there’s these are very deep topics now. But just knowing that their
477
00:31:32,958 –> 00:31:36,375
anthony_algmin: things is a great place to start. it’s a great place to understand. Hey,
478
00:31:36,708 –> 00:31:39,416
anthony_algmin: there is a lot of thought being put into these areas and there’s an
479
00:31:39,416 –> 00:31:42,791
anthony_algmin: evolution happening. I mean, it’s kind of crazy to me to talk a like ten
480
00:31:43,000 –> 00:31:47,750
anthony_algmin: years ago. These things were barely being figured out on the software
481
00:31:48,041 –> 00:31:51,500
anthony_algmin: engineering side, and and that’s true, it just seems like it’s been so long
482
00:31:51,750 –> 00:31:55,250
anthony_algmin: because they’ve become so ingrained in in our processes. But it’s it’s
483
00:31:55,416 –> 00:31:59,083
anthony_algmin: something too that you know they’ve evolved naturally and I think the data
484
00:31:59,250 –> 00:32:04,708
anthony_algmin: Baase has an interesting history because a lot of the analysis a lot of the
485
00:32:05,750 –> 00:32:09,083
anthony_algmin: understanding of what makes the data valuable and how it can be combined to
486
00:32:09,083 –> 00:32:11,750
anthony_algmin: create things that are relevant to the business happens on the business side
487
00:32:12,000 –> 00:32:15,166
anthony_algmin: itself, the data analyst, the financial analyst, Those are people that I
488
00:32:15,166 –> 00:32:20,000
anthony_algmin: wouldn’t call traditional I, t, or technology side of the house folks, but
489
00:32:20,541 –> 00:32:25,083
anthony_algmin: they, because of their need to leverage technology created these data groups
490
00:32:25,166 –> 00:32:28,958
anthony_algmin: that kind of came. Up in between that software engineering, sit on the very
491
00:32:29,083 –> 00:32:32,958
anthony_algmin: deep technical programming side and the data analyst, which were very
492
00:32:33,083 –> 00:32:35,916
anthony_algmin: business centr creating dashboards and things like that, and so you started
493
00:32:36,000 –> 00:32:41,333
anthony_algmin: to see to your point The the tooling started to cr, you know, create these
494
00:32:41,666 –> 00:32:46,958
anthony_algmin: these greater horizontal coverages that could be used by people that sat
495
00:32:47,083 –> 00:32:51,083
anthony_algmin: firmly on the business side of the house, but were able to exploit
496
00:32:51,083 –> 00:32:52,083
anthony_algmin: technology
497
00:32:52,708 –> 00:32:56,375
anthony_algmin: enhancements that went really deep into into the technical side of the
498
00:32:56,375 –> 00:33:00,041
anthony_algmin: house. So it’s it’s It’s an interesting paradigm that we find ourselves
499
00:33:00,208 –> 00:33:04,708
anthony_algmin: because to do this effectively, and and like what Maltono’s trying to do is
500
00:33:04,791 –> 00:33:08,625
anthony_algmin: to say, Okay, Well, you’re going to have a different collection of tools. Or
501
00:33:08,708 –> 00:33:11,250
anthony_algmin: what have you? We’re not going to try to replace that and I have a question
502
00:33:11,583 –> 00:33:14,625
anthony_algmin: related to that. But it’s it’s you’re You’re going to just try to plug in
503
00:33:14,791 –> 00:33:19,166
anthony_algmin: and explain it Because I think things like data lineage remain an incredibly
504
00:33:19,208 –> 00:33:20,208
anthony_algmin: difficult challenge
505
00:33:20,458 –> 00:33:23,833
anthony_algmin: because we’re looking at changes happening over time and that being able to
506
00:33:23,916 –> 00:33:28,125
anthony_algmin: retroactively go back and explore them. And we’ve already talked about how
507
00:33:28,375 –> 00:33:32,791
anthony_algmin: the development process of these data pipelines and andate movements we
508
00:33:32,875 –> 00:33:36,541
anthony_algmin: aren’t very good at showing our work and those by themselves, let alone
509
00:33:36,708 –> 00:33:40,041
anthony_algmin: looking back on it from the data perspective and saying where did this
510
00:33:40,125 –> 00:33:43,500
anthony_algmin: number come from? How did this get created? If we haven’t done our jobs in
511
00:33:43,666 –> 00:33:47,583
anthony_algmin: in documenting the flow of that that knowledge may be lost, let alone
512
00:33:47,833 –> 00:33:52,958
anthony_algmin: difficult to find. So the question that I have for you though is because we.
513
00:33:53,166 –> 00:33:56,125
anthony_algmin: we can certainly talk some more about lineage, and and that, because I think
514
00:33:56,208 –> 00:34:01,416
anthony_algmin: that’s a an area we’re talking about. But also how do you? for someone who’s
515
00:34:01,583 –> 00:34:05,166
anthony_algmin: in deep in the data spacease, that maybe doesn’t have as much of the that
516
00:34:05,250 –> 00:34:09,166
anthony_algmin: software engineering or programming background? how do you differentiate
517
00:34:09,500 –> 00:34:10,708
anthony_algmin: what Meltonno does
518
00:34:11,916 –> 00:34:17,500
anthony_algmin: versus being just another e t ▁l type of solution where you’re facilitating
519
00:34:17,666 –> 00:34:21,250
anthony_algmin: those workflows like, Help me draw that contrast between those Because I
520
00:34:21,333 –> 00:34:24,458
anthony_algmin: know there’s a relationship there, but I would understand better, the for
521
00:34:24,541 –> 00:34:27,833
anthony_algmin: the audience out there, I want them to understand better where does Meltonno
522
00:34:28,041 –> 00:34:34,125
anthony_algmin: sit in relation to what we would consider the e t ▁l um functionality that.
523
00:34:34,208 –> 00:34:35,666
anthony_algmin: We may be more accustomed to
524
00:34:36,041 –> 00:34:38,666
douwe_maan: Yeah, a great question. Theres a couple of things you were mentioned earlier.
525
00:34:38,833 –> 00:34:40,416
douwe_maan: I’d love to get back to one face that
526
00:34:40,500 –> 00:34:44,333
douwe_maan: kept coming up in my head while you were talking was just the importance of a
527
00:34:44,333 –> 00:34:48,750
douwe_maan: share vocabulary within data teams, but also between De and and software
528
00:34:48,916 –> 00:34:54,500
douwe_maan: engineering teams. But yeah, let’s talk a little bit first about Hamana relates
529
00:34:54,666 –> 00:35:01,208
douwe_maan: to T tos. For example, so um y ▁l t, you know e t, ▁l uh, whichever way you’re
530
00:35:01,291 –> 00:35:05,208
douwe_maan: going, data integration and transformation, uh is is, in many cases the first
531
00:35:05,291 –> 00:35:08,833
douwe_maan: step of the La. Data life cycle got to get the data from somewhere noted in some
532
00:35:08,916 –> 00:35:12,250
douwe_maan: data warehouse, and there’s a lot of solutions that have been built to do that,
533
00:35:12,416 –> 00:35:17,000
douwe_maan: Um. but we think they’re lacking in various ways. Um. they are not set up with
534
00:35:17,125 –> 00:35:21,083
douwe_maan: these des kind of best practices and principles. Mind, they don’t really allow
535
00:35:21,291 –> 00:35:25,458
douwe_maan: you to version your configuration or to test things locally before you. you
536
00:35:25,458 –> 00:35:29,458
douwe_maan: know. Um. have them take effect in the life environment. So one of the third
537
00:35:29,625 –> 00:35:30,916
douwe_maan: first things we started doing with
538
00:35:30,916 –> 00:35:32,041
douwe_maan: Miltano, and this is this
539
00:35:32,125 –> 00:35:35,000
douwe_maan: bivot about a year ago that I was talking about where I had to kind of figure
540
00:35:35,125 –> 00:35:39,458
douwe_maan: out how to turn the product around and find traction in the short term Was for
541
00:35:39,625 –> 00:35:43,541
douwe_maan: Melana which started with is really broad. One two for the entire entire data
542
00:35:43,708 –> 00:35:48,833
douwe_maan: life cycle. Uh, goal to focus on statches on that very first step and bring data
543
00:35:49,000 –> 00:35:52,916
douwe_maan: ups into it and and bring the atdtentages also of open source software into it.
544
00:35:53,083 –> 00:35:57,125
douwe_maan: So, Malana, if you go look at the websites, Um, you know in October, let’ saying
545
00:35:57,375 –> 00:36:00,166
douwe_maan: today, because we’ll have to see when this gets published. But as of October, on
546
00:36:00,250 –> 00:36:05,208
douwe_maan: the Montana website we position ourselves as E ▁l t, for the Data Os Era, and
547
00:36:05,625 –> 00:36:10,333
douwe_maan: Um, the way we compare to other E ▁l T tools. The open Tourt aspect makes a
548
00:36:10,333 –> 00:36:14,333
douwe_maan: really big difference, Um Melanno from They one has always been about I,
549
00:36:14,750 –> 00:36:19,625
douwe_maan: adopting and supporting existing, if possible, best in class open source
550
00:36:19,875 –> 00:36:24,833
douwe_maan: technologies and tools. So for E, ▁l, we chose Singer, which is a standard for
551
00:36:24,916 –> 00:36:28,166
douwe_maan: data connectors that has a library currenturly of about three hundred connectors
552
00:36:28,333 –> 00:36:31,791
douwe_maan: for different sources and destinations. Um. In that sense, it can rival the
553
00:36:31,791 –> 00:36:35,208
douwe_maan: libraries of tools like Five trend, but this is completely open source. All the
554
00:36:35,208 –> 00:36:38,666
douwe_maan: source codes is available. Anyone in the world can build a new connector with E,
555
00:36:38,833 –> 00:36:42,041
douwe_maan: and we build an s. T case or for development kit to make that extremely easy,
556
00:36:42,583 –> 00:36:48,250
douwe_maan: and Um. People can improve extent iterate all these existent connectors as they
557
00:36:48,333 –> 00:36:51,791
douwe_maan: see fits to make them more appropriate for their use case. so when it’s just
558
00:36:51,958 –> 00:36:55,958
douwe_maan: about e t, ▁l Motanno has massive advantages over other solution because of its
559
00:36:56,041 –> 00:37:00,125
douwe_maan: much larger library of connectors and the flexibility it gives teams to go
560
00:37:00,250 –> 00:37:04,750
douwe_maan: beyond for the support out of the box, Um. And as an open source platform, it
561
00:37:04,916 –> 00:37:09,291
douwe_maan: can be self managed that you have complete control over uh, where the data goes
562
00:37:09,458 –> 00:37:13,541
douwe_maan: and where it doesn’t go, perhaps more importantly of for privacy security, or
563
00:37:13,625 –> 00:37:16,583
douwe_maan: even compliance reasons, if you’re working with healthcare data, for example,
564
00:37:16,583 –> 00:37:17,583
anthony_algmin: Hey
565
00:37:17,583 –> 00:37:21,958
douwe_maan: But for Meltao’s perspective, Singer for data connectors and d, b, t, uh. This,
566
00:37:22,125 –> 00:37:24,916
douwe_maan: this amazing Da of transformation told than I’m sure most businessers are
567
00:37:25,000 –> 00:37:29,208
douwe_maan: familiar with are just a couple of the plug ins that our Daate out platform
568
00:37:29,541 –> 00:37:32,666
douwe_maan: supports, and the goal of Melannos to provide that platform that you can bring
569
00:37:32,833 –> 00:37:36,500
douwe_maan: different components into, and then Melana provides the glued. The intermediate
570
00:37:36,583 –> 00:37:41,000
douwe_maan: tissue between the different components gives you a uh, consistent configuration
571
00:37:41,125 –> 00:37:45,708
douwe_maan: layer for the entire the thing, and also allows the different components to Um,
572
00:37:45,958 –> 00:37:49,375
douwe_maan: have their configurations synchronized automatically so that when you point at a
573
00:37:49,375 –> 00:37:52,750
douwe_maan: different snowfake warehouse, all the tools, including potentially, your, you
574
00:37:52,750 –> 00:37:58,250
douwe_maan: know, your B, I or analysis tool knows where to look now. Um, So for E ▁l T
575
00:37:58,333 –> 00:38:02,333
douwe_maan: alone, the combination of Singer and d B T managed with Mel Anno, can directly
576
00:38:02,500 –> 00:38:05,541
douwe_maan: compete with five trainn, or all these other tools people might be using today.
577
00:38:05,791 –> 00:38:09,708
douwe_maan: But what Montana really adds is that layer that brings together great open
578
00:38:09,875 –> 00:38:13,000
douwe_maan: source tools and technologies. So when you’re thinking about your modern, your
579
00:38:13,125 –> 00:38:17,375
douwe_maan: datateck today, Uh, and you have five tooling decisions. Some of them, you might
580
00:38:17,458 –> 00:38:20,833
douwe_maan: have gone for some proprietary Sa software, but there exist really great
581
00:38:21,083 –> 00:38:26,041
douwe_maan: open source. Uh, the itols like supers sett metaabase today as well, Um, you
582
00:38:26,125 –> 00:38:30,666
douwe_maan: know, on the data orchestration, Uh, side of things, for example, when you are
583
00:38:30,833 –> 00:38:35,000
douwe_maan: talking about these workflow managers like Airflow and Dexter and Prefect, this
584
00:38:35,083 –> 00:38:37,958
douwe_maan: is another piece of software that’s open source and you mighting into the place
585
00:38:38,166 –> 00:38:41,625
douwe_maan: somewhere, Um, and then if you want to have d v T running somewhere with your
586
00:38:41,708 –> 00:38:45,291
douwe_maan: transformation pipeps, then again, that’s another piece to deploy, and this
587
00:38:45,541 –> 00:38:48,333
douwe_maan: become this becomes really complicated very quickly because you need all of
588
00:38:48,333 –> 00:38:49,333
douwe_maan: these tools to
589
00:38:49,333 –> 00:38:52,666
douwe_maan: have some shared understanding of what. We’re all about, and it is Meltano that
590
00:38:52,750 –> 00:38:55,958
douwe_maan: provides that stable foundation on top of which you and built your ideal
591
00:38:56,041 –> 00:39:00,750
douwe_maan: datastack, And Malano allows teams to treat all of this as if it is one product
592
00:39:01,000 –> 00:39:05,000
douwe_maan: that they build with one repository where they can do this version, control and
593
00:39:05,083 –> 00:39:08,750
douwe_maan: coach review work. And it allows entire data teams, all the different
594
00:39:09,000 –> 00:39:12,500
douwe_maan: disciplines that might exist from data engineers, andliytics, engineers and
595
00:39:12,583 –> 00:39:16,625
douwe_maan: analysts and scientists to come together and have a single source of truth in a
596
00:39:16,625 –> 00:39:17,625
douwe_maan: single place
597
00:39:17,708 –> 00:39:22,041
douwe_maan: to discuss and collaborate and give each other feedback on anyone’s idea for how
598
00:39:22,166 –> 00:39:26,166
douwe_maan: this data product could be better. Um, kind of to tiit us back to what you were
599
00:39:26,250 –> 00:39:29,375
douwe_maan: saying about the share, or what I kind of heard you say about the shared
600
00:39:29,458 –> 00:39:34,333
douwe_maan: forcovery and the importance of it just by bringing people together in one
601
00:39:34,916 –> 00:39:39,083
douwe_maan: place. One website that defines the entirety of the data products all way from
602
00:39:39,125 –> 00:39:44,500
douwe_maan: integration to Uh. analysis, allows them to ask each other questions and make
603
00:39:44,583 –> 00:39:47,958
douwe_maan: suggestions and go find the relevant code and see like Hey, is this where I
604
00:39:48,041 –> 00:39:50,750
douwe_maan: should make a change if I want ▁x. that. Instead of saying well, I’m a date
605
00:39:50,916 –> 00:39:54,500
douwe_maan: analyst, I have no idea about this engineering stuff if I need anything, even if
606
00:39:54,583 –> 00:39:57,708
douwe_maan: it’s just another column or another, a p. i, N point that I would love to see in
607
00:39:57,791 –> 00:40:01,375
douwe_maan: my database. I need to create a ticket and throw through the data engineers, and
608
00:40:01,458 –> 00:40:05,000
douwe_maan: then they are going to be working in this kind of like ticket response workflow,
609
00:40:05,125 –> 00:40:08,333
douwe_maan: just like analysts often do when they just get questions from the business. And
610
00:40:08,416 –> 00:40:11,625
douwe_maan: and they basically its working through tickets all day, But it would be far more
611
00:40:11,875 –> 00:40:15,208
douwe_maan: effective for all of these stakeholders to be able to have a shaft for caavalry.
612
00:40:15,458 –> 00:40:19,083
douwe_maan: be able to trace. Okay, I’m looking at this dashboard here. I can find the code
613
00:40:19,125 –> 00:40:22,583
douwe_maan: that pulled that data from the rel of a B. I. If I’m a little bit familiar with
614
00:40:22,583 –> 00:40:25,708
douwe_maan: a sequel, or maybe a little bit of Python or Yammo files, where you can
615
00:40:25,791 –> 00:40:29,208
douwe_maan: configure the stuff, I can figure out how to add another attribute. And because
616
00:40:29,458 –> 00:40:32,416
douwe_maan: this is not just immediately going to go life, but it’s going to go through this
617
00:40:32,500 –> 00:40:36,125
douwe_maan: Sce review process. You can have the confidence to propose changes even if
618
00:40:36,125 –> 00:40:37,458
douwe_maan: you’re not super confident what
619
00:40:37,458 –> 00:40:40,416
douwe_maan: you’re doing because you know it’s going to be reviewed and approved by someone
620
00:40:40,583 –> 00:40:43,125
douwe_maan: who knows better, and not just a person who knows better with an actual
621
00:40:43,291 –> 00:40:45,708
douwe_maan: automated testing pipeline. And it will tell you if you
622
00:40:45,791 –> 00:40:51,625
douwe_maan: accidentally broke something or not before. Um, you know, as an alternative to
623
00:40:51,958 –> 00:40:54,916
douwe_maan: saying, I don’t even want to touch this, because I’ll break things. You guys do
624
00:40:55,000 –> 00:40:56,583
douwe_maan: it and free
625
00:40:56,833 –> 00:40:59,708
douwe_maan: these people closer together. Not just on the data, Cyle on the soffign,
626
00:40:59,875 –> 00:41:04,125
douwe_maan: Engineerys is is extremely valuable and a really big part of trying to
627
00:41:04,125 –> 00:41:06,916
douwe_maan: accomplish and and what theffs has accomplished. And so for development.
628
00:41:08,708 –> 00:41:12,291
anthony_algmin: Well in and it. it encourages what we were talking about earlier around.
629
00:41:12,625 –> 00:41:17,583
anthony_algmin: learning more on the other side of the coin is is that if you can feel free
630
00:41:17,916 –> 00:41:24,000
anthony_algmin: and encouraged to suggest potential solutions, I think of it as a as a great
631
00:41:24,208 –> 00:41:28,625
anthony_algmin: way on the technology side of things, to do what I do in in business and
632
00:41:28,875 –> 00:41:33,916
anthony_algmin: consulting all the time. It’s to say hey, half of consulting is suggesting
633
00:41:34,125 –> 00:41:38,458
anthony_algmin: really bad ideas for people to react to and and going from them. just
634
00:41:38,541 –> 00:41:42,375
anthony_algmin: because of a blank piece of paper, is a much more difficult starting place
635
00:41:42,791 –> 00:41:46,708
anthony_algmin: than a terrible idea. Like a terrible idea You can react to very quickly and
636
00:41:46,791 –> 00:41:51,000
anthony_algmin: iterate fast and you’ll get somewhere Uh, remarkably quick. If I know I’m
637
00:41:51,083 –> 00:41:53,916
anthony_algmin: looking for something and I can say Hey, it should look something like this,
638
00:41:54,375 –> 00:41:58,708
anthony_algmin: And then there’s a way for me to propose that, and for the system or for the
639
00:41:58,791 –> 00:42:01,583
anthony_algmin: people to react to it and make it. Oh, you know what we should do it this
640
00:42:01,666 –> 00:42:06,541
anthony_algmin: way Because of ▁x, You could be so much closer, so much faster with that
641
00:42:06,708 –> 00:42:12,000
anthony_algmin: kind of interaction and it helps self document. Like, if I can just show you
642
00:42:12,125 –> 00:42:16,875
anthony_algmin: what I’m talking about and then we can react, It saves me the trouble of
643
00:42:17,000 –> 00:42:20,791
anthony_algmin: trying to figure out the words to describe the thing that I could otherwise
644
00:42:20,958 –> 00:42:24,625
anthony_algmin: draw you a more direct picture of. And And that’s the kind of. I think
645
00:42:24,708 –> 00:42:28,875
anthony_algmin: that’s where Dev Ops is is at its finest right where we collectively bring
646
00:42:29,083 –> 00:42:34,208
anthony_algmin: our wisdom to a shared. You know it’s it’s common vocabulary, C, common
647
00:42:34,541 –> 00:42:39,833
anthony_algmin: vocabulary, but it’s also a common working area where we together solve
648
00:42:40,041 –> 00:42:43,166
anthony_algmin: these problems using our different perspectives, our differentvantage
649
00:42:43,333 –> 00:42:48,041
anthony_algmin: points, or different skill sets, to then collectively solve whatever it is
650
00:42:48,208 –> 00:42:52,000
anthony_algmin: that needs to be done. Get it done in the fewest steps necessary.
651
00:42:53,000 –> 00:42:58,458
anthony_algmin: have it be validated and then codified Put in production. You’re done. You
652
00:42:58,541 –> 00:43:01,333
anthony_algmin: move on and and you get to the next thing that much faster, And that to me
653
00:43:01,666 –> 00:43:06,041
anthony_algmin: is is the key takeaway for folks that are are new to the space. is to say,
654
00:43:06,875 –> 00:43:14,000
anthony_algmin: this is all about evolving how we work to get to the core essence of what’s
655
00:43:14,125 –> 00:43:19,000
anthony_algmin: really necessary, and remove as much of the redundancy as much of the than
656
00:43:19,250 –> 00:43:23,833
anthony_algmin: unnecessary steps as possible. And this is a continued evolution, right,
657
00:43:24,208 –> 00:43:28,375
anthony_algmin: this is, we’ve got. We’ve come a long way in the last ten years, Right.
658
00:43:28,708 –> 00:43:30,375
anthony_algmin: What? What are these next ten years?
659
00:43:31,500 –> 00:43:34,458
anthony_algmin: What will they have in stores? So that that’s my next question for you, Like
660
00:43:34,625 –> 00:43:38,208
anthony_algmin: where do you see this going? What’s the natural evolution of this And and
661
00:43:38,291 –> 00:43:42,000
anthony_algmin: how does Meltano Um expect to play a role in that?
662
00:43:42,916 –> 00:43:47,291
douwe_maan: Yeah, yeah, left. to answer that question. I also love to start by kind of tying
663
00:43:47,458 –> 00:43:51,000
douwe_maan: into what you said about the importance of the common working area, and that’s
664
00:43:51,083 –> 00:43:54,833
douwe_maan: really what I’m talking about when we say you know one repository, which is kind
665
00:43:54,916 –> 00:43:58,250
douwe_maan: of a software development way of saying you, a project, a collection of files
666
00:43:58,416 –> 00:44:02,041
douwe_maan: that come together to create something, a software product, in many cases, or
667
00:44:02,166 –> 00:44:06,125
douwe_maan: the single source of truth, where you give the entire team one place where they
668
00:44:06,166 –> 00:44:10,833
douwe_maan: know they can find the actual state of things, and also knowing that everything
669
00:44:11,208 –> 00:44:15,000
douwe_maan: that comes together to create that. Dashboard’re looking at a Yourbi tool is in
670
00:44:15,083 –> 00:44:18,916
douwe_maan: one place, and you could theoretically, if you as an analyst, detected, or some
671
00:44:19,083 –> 00:44:25,625
douwe_maan: Vaian database, use Melttao to um. find out within one place. the code that was
672
00:44:25,708 –> 00:44:29,375
douwe_maan: responsible for pulling out of the api. The code was responsible for from
673
00:44:29,625 –> 00:44:34,125
douwe_maan: transforming it into the required data. Schma, tools that were responsible for
674
00:44:34,500 –> 00:44:38,583
douwe_maan: you know, having anomeditated validation tests while the pipeline was running
675
00:44:38,750 –> 00:44:42,750
douwe_maan: all the way to the actual sequel Creer that generates report, giving people the
676
00:44:42,833 –> 00:44:45,791
douwe_maan: confidence that they know where to find it. Instead of this scary feeling of
677
00:44:45,875 –> 00:44:50,041
douwe_maan: like, Well, I’m an analyst. I know byb I tool, but anything that comes before it
678
00:44:50,125 –> 00:44:53,708
douwe_maan: is weird, tools never used and none of the accounts. I don’t know which buttons
679
00:44:53,791 –> 00:44:57,625
douwe_maan: to click. I’m afraid to break things. Um, giving them the confidence they can
680
00:44:57,625 –> 00:44:59,458
douwe_maan: find it, and like you said, that they can suggest the change
681
00:44:59,625 –> 00:45:03,375
douwe_maan: no matter how bad it is, and not accidentally break things is massive and in
682
00:45:03,458 –> 00:45:07,291
douwe_maan: Soware development, what we saw is in just by bringing product managers and U, y
683
00:45:07,458 –> 00:45:11,958
douwe_maan: u, Y, designers and software engineers. And and you know, K, a quality testers
684
00:45:12,166 –> 00:45:17,083
douwe_maan: to one place and giving them um access to everything the others are seeing,
685
00:45:17,541 –> 00:45:21,875
douwe_maan: allows this collaboration to turn from a kind of walled system where you just
686
00:45:21,875 –> 00:45:25,208
douwe_maan: throw stuff over fences because you’re afraid to even touch it. Because you know
687
00:45:25,291 –> 00:45:28,833
douwe_maan: you could break things. That’s kind of the reality of things today. Um, we want
688
00:45:28,916 –> 00:45:32,041
douwe_maan: to throw that on its head. So, when we’re talking about where the next ten years
689
00:45:32,166 –> 00:45:36,166
douwe_maan: are going, like I was saying earlier, Um, in data, we’re still at the space now
690
00:45:36,333 –> 00:45:40,125
douwe_maan: where it’s not even so much about convincing people that Nottana is amazing.
691
00:45:40,250 –> 00:45:43,208
douwe_maan: It’s about convincing that the data Os is something they want at all. That
692
00:45:43,875 –> 00:45:47,458
douwe_maan: deas to data makes sense that software development teams have a lot more in
693
00:45:47,541 –> 00:45:51,708
douwe_maan: common with data teams than they might realize. Um. some of this kind of wave
694
00:45:52,250 –> 00:45:57,791
douwe_maan: towards a horizontally integrated data tooling space with besting class tools.
695
00:45:58,250 –> 00:46:03,083
douwe_maan: Um, is already kind of revolutionizing the way things work. but something has
696
00:46:03,208 –> 00:46:07,291
douwe_maan: gone missing because there’s no longer that that single place people know to go.
697
00:46:07,875 –> 00:46:12,500
douwe_maan: So we, in the next ten years are going to be seeing a combination of I think,
698
00:46:13,208 –> 00:46:18,041
douwe_maan: really, Um. Focused tools for really specific problems, because there can be ten
699
00:46:18,166 –> 00:46:21,291
douwe_maan: competitors and everyone can kind of bring in their new idea to make
700
00:46:21,458 –> 00:46:25,083
douwe_maan: transformation better than ever was before, while trying to
701
00:46:25,125 –> 00:46:29,541
douwe_maan: fit this into a world where there’s clearly still value in having a consistent
702
00:46:29,875 –> 00:46:32,916
douwe_maan: foundation. That that pulls it all together. Of course that’s exactly what
703
00:46:33,000 –> 00:46:37,083
douwe_maan: Moltana is, but this is a challenge that will take years and years to kind of
704
00:46:37,208 –> 00:46:41,125
douwe_maan: roll out in in many organizations, and and some of this will also just have to
705
00:46:41,375 –> 00:46:45,875
douwe_maan: become companies getting comfortable getting off of their maybe ten year old an
706
00:46:46,041 –> 00:46:50,416
douwe_maan: to end data platforms, adopting specific tools for each step of the way in
707
00:46:50,500 –> 00:46:55,958
douwe_maan: combination with it to like moanno, and data teams that need to not sly, retrain
708
00:46:56,125 –> 00:46:58,500
douwe_maan: their people, because if these people are productive today than they can ever
709
00:46:58,583 –> 00:47:01,875
douwe_maan: be, but at least realizing there’s additional skills. They can learn and there
710
00:47:02,041 –> 00:47:05,208
douwe_maan: are things from software development that are really not specific to software
711
00:47:05,375 –> 00:47:09,625
douwe_maan: development. Um. Get Lab. Their mission do, mission for Getlab has always been,
712
00:47:09,958 –> 00:47:12,166
douwe_maan: everyone can contribute, and the the
713
00:47:12,250 –> 00:47:15,708
douwe_maan: broader vision has been to build a platform that allows knowledge workers in any
714
00:47:15,875 –> 00:47:19,875
douwe_maan: profession to more effectively coliaborate to build better things together. And
715
00:47:20,041 –> 00:47:23,458
douwe_maan: in this mission division software is explicitly mentioned nowhere. like
716
00:47:23,541 –> 00:47:27,125
douwe_maan: obviously today Getit Lab is still primarily focused in software development
717
00:47:27,291 –> 00:47:31,791
douwe_maan: teams E. but now Getlab and Devts in general, is it disposition where it can
718
00:47:31,916 –> 00:47:32,916
douwe_maan: start moving
719
00:47:33,000 –> 00:47:36,041
douwe_maan: into more and more and more industries, starting with the ones that are more
720
00:47:36,166 –> 00:47:40,500
douwe_maan: closely aligned more adjacent to software development. But the goal is very much
721
00:47:40,750 –> 00:47:44,833
douwe_maan: to find a way to bring it advantages to any knowledge industry in the world, and
722
00:47:45,125 –> 00:47:49,208
douwe_maan: part of the challenge with Moltanno is to kind of provide that layer between the
723
00:47:49,291 –> 00:47:52,666
douwe_maan: software development way of looking at things and the data away of looking at
724
00:47:52,833 –> 00:47:58,750
douwe_maan: things, Um, and and doing it in such a way that the Defap’s best practices will
725
00:47:58,833 –> 00:48:03,875
douwe_maan: feel natural to data teams, and in the same sense that now data work will feel
726
00:48:04,041 –> 00:48:08,125
douwe_maan: natural to software development teams, so over the, that’s part of what I see
727
00:48:08,250 –> 00:48:11,375
douwe_maan: happening this is. Really big data of revolution, which is going to take ten
728
00:48:11,541 –> 00:48:15,875
douwe_maan: years to pen out, Um. The other thing that I think is really key is open source
729
00:48:16,125 –> 00:48:17,458
douwe_maan: software. One of the many
730
00:48:17,625 –> 00:48:22,833
douwe_maan: reasons why meltaa’ open source um is because I just very strongly personally
731
00:48:23,083 –> 00:48:29,375
douwe_maan: believe that the most, the best most useful and most effective tool will always
732
00:48:29,541 –> 00:48:33,708
douwe_maan: be a tool built in very close collaboration with the users of that tool. The
733
00:48:33,708 –> 00:48:37,458
douwe_maan: feedback loop needs to be incredibly tight, and it does not get tighter than
734
00:48:37,625 –> 00:48:41,791
douwe_maan: inviting your users into your issue, Tcker into your code base, allowing them to
735
00:48:41,791 –> 00:48:45,208
douwe_maan: make suggestions like you, said Anthony. Allowment makes suggestions. even if
736
00:48:45,208 –> 00:48:48,333
douwe_maan: they will be. the code will not be up to par or whatever, which is. giving that
737
00:48:48,500 –> 00:48:52,750
douwe_maan: ability to express themselves and start a conversation. Allows us to build tools
738
00:48:52,916 –> 00:48:57,791
douwe_maan: that make its user more effective than ever, and in Getlab itself it was always
739
00:48:57,875 –> 00:49:00,583
douwe_maan: quite special that the users of the product for software engineers, and
740
00:49:00,583 –> 00:49:03,458
douwe_maan: obviously the people building the protocols software engineers. But what
741
00:49:03,541 –> 00:49:07,541
douwe_maan: foreseeing in data now similarly is that Data engineers and Olympics engineers
742
00:49:07,625 –> 00:49:11,000
douwe_maan: are increasingly becoming more confident with little bits of programming and a
743
00:49:11,000 –> 00:49:15,083
douwe_maan: little bits of Python, and having the opportunity to build your own ideal tool
744
00:49:15,208 –> 00:49:19,458
douwe_maan: to make you most productive. Um, it doesn’t get better than that, and we can
745
00:49:19,541 –> 00:49:23,458
douwe_maan: bring together the ideas and the perspectives of of thousands, if not a millions
746
00:49:23,541 –> 00:49:27,625
douwe_maan: of data engineers, instead of having to defer to one product manager and their
747
00:49:27,791 –> 00:49:32,125
douwe_maan: team. However amazing they might be, It’s different from saying this is a
748
00:49:32,250 –> 00:49:35,291
douwe_maan: product. Four data people by data people, Um,
749
00:49:36,416 –> 00:49:41,083
douwe_maan: our mission very explicitly is to not force our perspective on you, especially
750
00:49:41,208 –> 00:49:44,583
douwe_maan: because I coming from software development. Never having worked in data are very
751
00:49:44,750 –> 00:49:49,291
douwe_maan: aware of how little I know, but I really want to listen and create a space for
752
00:49:49,541 –> 00:49:53,000
douwe_maan: people to come together and build really awesome data to link together. And I
753
00:49:53,083 –> 00:49:58,583
douwe_maan: think that in any field, any industry where the users of tools have the
754
00:49:58,750 –> 00:50:02,583
douwe_maan: capability of programming even if it’s just as little as no. I don’ know how to
755
00:50:02,583 –> 00:50:07,000
douwe_maan: write simple code in an editor, and I’ so beginning to use Get. Once you have
756
00:50:07,208 –> 00:50:11,291
douwe_maan: that level of technical skills in your users, open source software will always
757
00:50:11,458 –> 00:50:16,666
douwe_maan: win. period, because this model of we chargear for it, we you know, build it to
758
00:50:16,666 –> 00:50:20,333
douwe_maan: our own liking. If you don’t like it, go to a competitor just does not skill. If
759
00:50:20,416 –> 00:50:24,166
douwe_maan: you have a user base that is empowered to say, Oh, we’re fed up’re, just going
760
00:50:24,166 –> 00:50:26,500
douwe_maan: to buil something yourself, and it’s going to be better because they can
761
00:50:26,666 –> 00:50:31,000
douwe_maan: actually make it happen. So, Um, more and more, I think that that, um, the
762
00:50:31,208 –> 00:50:34,583
douwe_maan: entire space will move to open source software. To some extent we’re already
763
00:50:34,750 –> 00:50:38,416
douwe_maan: seeing that. of course, with Um, you know, work for orchanation, stuff like
764
00:50:38,500 –> 00:50:43,125
douwe_maan: Airflo Exxtra and Prefect d B T for transformation, Uh, and especially now that
765
00:50:43,291 –> 00:50:47,000
douwe_maan: the space is becoming increasingly horizontally integrated, it’s going to be
766
00:50:47,000 –> 00:50:49,875
douwe_maan: much easier for anyone to step up and say Well, I built an open tource
767
00:50:50,041 –> 00:50:54,250
douwe_maan: alternative to fancy expensive tool, ▁x, And then that’s open tour. Our tool
768
00:50:54,416 –> 00:50:57,958
douwe_maan: might actually be much better than the alternative places, especially if it can
769
00:50:58,041 –> 00:51:02,041
douwe_maan: rally around another couple hundred people to bring in their ideas, perspectives
770
00:51:02,125 –> 00:51:07,125
douwe_maan: and en coding skills. So ten years from now, just like we see in software
771
00:51:07,291 –> 00:51:10,583
douwe_maan: development, and the database is going to be built around open force. Sting.
772
00:51:10,833 –> 00:51:15,541
douwe_maan: Devil practices are going to be built in from day one, and these data teams are
773
00:51:15,625 –> 00:51:19,000
douwe_maan: going to be functioning and thinking of themselves, much more like teams
774
00:51:19,125 –> 00:51:23,791
douwe_maan: building a product. Then you know I, I work in two. A you working to be working
775
00:51:23,875 –> 00:51:27,083
douwe_maan: to C, and the responsibility of the data team is to make good product of
776
00:51:27,083 –> 00:51:30,250
douwe_maan: purchasing decisions, which I think is a really kind of sad way of thinking
777
00:51:30,416 –> 00:51:34,750
douwe_maan: about this. this massive responsibility of uh, setting up all the infrastructure
778
00:51:34,916 –> 00:51:39,541
douwe_maan: a company needs to use data to the full potential and tying together an open
779
00:51:39,708 –> 00:51:43,208
douwe_maan: source and the full potential. Bits. Uh, you, you end up at what our mission is
780
00:51:43,291 –> 00:51:47,375
douwe_maan: as Mofana, which is to enable everyone to realize the full potential of their
781
00:51:47,458 –> 00:51:53,000
douwe_maan: data And everyone here is intentionally broad and we want to include anyone who
782
00:51:53,083 –> 00:51:55,958
douwe_maan: could benefit from doing anything with their data which we know is literally
783
00:51:56,125 –> 00:52:00,583
douwe_maan: everyone, and give them the ability to do so in an open source, Ties to that in
784
00:52:00,583 –> 00:52:01,583
douwe_maan: a really big way.
785
00:52:03,666 –> 00:52:07,750
anthony_algmin: thank you so much for sharing that. Unfortunately we’re out of time. I’m I.
786
00:52:07,750 –> 00:52:12,958
anthony_algmin: I’m so excited about what Meltono is doing about how you’re bringing open
787
00:52:13,250 –> 00:52:15,583
anthony_algmin: source and data opts to everybody.
788
00:52:16,875 –> 00:52:21,083
anthony_algmin: I. I’m thrilled with your business model. I think this is amazing stuff and
789
00:52:21,166 –> 00:52:27,083
anthony_algmin: I think it’s really going to help drive that next evolution of what kind of
790
00:52:27,250 –> 00:52:32,458
anthony_algmin: open source and and project management and the way we collaborate in the Dat
791
00:52:32,625 –> 00:52:36,208
anthony_algmin: spacease. I think it’s It’s overdue and and it’s great to see you putting
792
00:52:36,375 –> 00:52:40,291
anthony_algmin: some real energy behind that. I can’t wait to see what’s next for you in
793
00:52:40,375 –> 00:52:44,458
anthony_algmin: Milta, and to watch this growth over coming years. I think it’s going to be
794
00:52:44,541 –> 00:52:47,750
anthony_algmin: an amazing ride. So thank you for being on the show and sharing some of that
795
00:52:48,250 –> 00:52:51,791
douwe_maan: Yeah, thank you, Anthony so much for the opportunity and for everyone Listening.
796
00:52:51,958 –> 00:52:56,416
douwe_maan: If this story of data people building tools for data people resonates, please
797
00:52:56,666 –> 00:53:00,500
douwe_maan: come join our Slect community where we currently have over eighteen hundred data
798
00:53:00,666 –> 00:53:03,625
douwe_maan: professionals. Uh, doing exactly what I’ve been describing coming together,
799
00:53:03,875 –> 00:53:06,833
douwe_maan: sharing ideas and perspectives and trying to build the next generation of
800
00:53:06,833 –> 00:53:07,833
douwe_maan: dataing,
801
00:53:08,500 –> 00:53:10,666
douwe_maan: and you’re more welcome to come join the party.
802
00:53:11,916 –> 00:53:12,500
anthony_algmin: And we will definitely include Uh, more information and links uh to that in
803
00:53:12,500 –> 00:53:17,333
anthony_algmin: And we will definitely include Uh, more information and links uh to that in
804
00:53:17,416 –> 00:53:21,166
anthony_algmin: the show note. So so check out the show notes for Uh directions on how to
805
00:53:21,166 –> 00:53:25,000
anthony_algmin: get to Meltano and and where you can a join that community. So thank you all.
806
00:53:25,250 –> 00:53:29,000
anthony_algmin: Uh for joining us today. Uh, dive deeper with my book at Data Leadership
807
00:53:29,166 –> 00:53:33,416
anthony_algmin: Book Dot Com and use Promocode Almondel at the Diversity Online Trainding
808
00:53:33,500 –> 00:53:37,000
anthony_algmin: Center for twenty percent off your first purchase. and if you enjoy our show
809
00:53:37,166 –> 00:53:40,208
anthony_algmin: and would love your own, but don’t know where to start, visit Algmin.com
810
00:53:40,291 –> 00:53:44,000
anthony_algmin: to learn how we make having your own video podcast as easy as joining a
811
00:53:44,000 –> 00:53:47,916
anthony_algmin: call and sending an email. Stay safe during these unusual times and go make
812
00:53:47,916 –> 00:53:48,916
anthony_algmin: an impact
