{"id":1823,"date":"2025-11-04T23:52:57","date_gmt":"2025-11-05T00:52:57","guid":{"rendered":"http:\/\/gogetmuscle.com\/?p=1823"},"modified":"2025-11-05T17:47:11","modified_gmt":"2025-11-05T17:47:11","slug":"gtm-169-how-airbyte-hit-1b-the-open-source-community-first-playbook","status":"publish","type":"post","link":"http:\/\/gogetmuscle.com\/index.php\/2025\/11\/04\/gtm-169-how-airbyte-hit-1b-the-open-source-community-first-playbook\/","title":{"rendered":"GTM 169: How Airbyte Hit $1B: The Open-Source, Community-First Playbook"},"content":{"rendered":"
The GTM Podcast is available on any major directory, including:<\/p>\n
Michel Tricot<\/a> is the co-founder and CEO of Airbyte, the open-source data movement platform he launched in 2020. Before Airbyte, Michel led integrations and served as Director of Engineering at LiveRamp, where he scaled the teams and pipelines that synced massive data volumes. He also helped build rideOS as a founding engineer and Director of Engineering. Michel has spent 15+ years in data infrastructure, with a focus on commoditizing data pipelines and giving teams control and sovereignty over their data.<\/p>\n 00:15 \u2014 Airbyte\u2019s rise: open source, community-first, and a billion\u2011plus valuation. 01:57 \u2014 Michel explains Airbyte in two lines: open data movement into warehouses (and now agents). 02:53 \u2014 Why launch open source on GitHub: capture engineers at the \u201cwrite a painful script\u201d moment. 06:53 \u2014 COVID reset: from a marketing\u2011focused product to an OSS platform that hit a hockey\u2011stick curve. 11:01 \u2014 Project-market fit vs product-market fit: adoption is not monetization. 14:41 \u2014 How Airbyte turned Slack into a rapid product feedback loop (ship next\u2011day fixes). 19:22 \u2014 The community trap: when your Slack becomes support, and how they course\u2011corrected. 23:53 \u2014 Cloud the hard way: why customers wanted control\/sovereignty more than a hosted version. 29:22 \u2014 Building an enterprise motion: hire earlier, expect 6\u20139 month ramps, many more stakeholders. 33:26 \u2014 Fast path to Series A: publishing the deck, OSS adoption surge, and choosing investor fit. 1. Shrink scope to find signal.<\/strong> 2. Separate project-market fit from product-market fit.<\/strong> 3. Ship transparency as a growth channel.<\/strong> 4. Community needs design, not just support.<\/strong> 5. Control beats convenience in data infra.<\/strong> 6. Don\u2019t hire ahead of platform complexity.<\/strong> 7. Content compounds when founder-led.<\/strong> 8. Use community for real-time product discovery.<\/strong> 9. Enterprise motion is human-time, not server-time.<\/strong> 10. Build for agents, not just analysts.<\/strong> ZoomInfo is the GTM Intelligence Platform built for sales, marketing, and RevOps.<\/p>\n By unifying data, workflows, and insights into a single system, ZoomInfo helps revenue teams find and engage the right buyers, launch go-to-market plays faster, and drive predictable growth. With industry-leading accuracy and depth of data, it gives your team the intelligence advantage to win in competitive markets.<\/p>\n It\u2019s trusted by the fastest-growing companies and has become the category leader in GTM Intelligence.<\/p>\n Learn more at zoominfo.com<\/a>.<\/p>\n Michel Tricot:\u00a0<\/strong>0:00<\/p>\n People are willing to put time into the project and the product that we are building. How do you actually commercialize it? It\u2019s a different story. And to me, that\u2019s what PMF actually is, where everything goes super fast, every deal gets closed in like a week, two weeks, one month max.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>0:15<\/p>\n You launched in 2020. Now you\u2019re valued at over a billion dollars. Michelle didn\u2019t play the typical fast launch playbook. He went open source first, community first, and GitHub first. And it ignited one of the fastest bottoms-up adoption curves in modern data infrastructure. Today, AirPyte is valued at over a billion dollars, powering data movements for thousands of teams, including over 20% of the Fortune 500. And they got there in a really interesting way. They built in public, compounded through community, and turned contribution into a distribution mode. In this conversation, we break down the stories and lessons behind all of this growth, including a really important lesson on separating product market fit from project market fit. All right, let\u2019s get into it. Michelle, welcome to the podcast.<\/p>\n Michel Tricot:\u00a0<\/strong>1:11<\/p>\n Thank you for having me. Great to be there.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>1:13<\/p>\n It is a pleasure. And it hasn\u2019t been long since we saw each other earlier this week, in fact. So it\u2019s great to see you again.<\/p>\n Michel Tricot:\u00a0<\/strong>1:19<\/p>\n Yeah, no, that was a good, a good event. Like that was the Tech Crunch one that was very solid.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>1:25<\/p>\n Yeah, that was great. And your your uh session was highly attended. Sounded fantastic. Excited to pick your brain a little bit more intimately than at the event itself. And you launched in 2020. Now you\u2019re valued at over a billion dollars. Take us back. We want to know the how behind this type of growth.<\/p>\n Michel Tricot:\u00a0<\/strong>1:43<\/p>\n Yeah.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>1:44<\/p>\n And before we even get started from the beginning, a lot of our audience aren\u2019t engineers. A lot of operators, a lot of founders. Give us a high level of AirBite. What does Airbite do? Kind of the two-liner for everyone listening.<\/p>\n Michel Tricot:\u00a0<\/strong>1:57<\/p>\n Yeah. So AirBite is an open data movement platform, meaning that we can take any pieces of data across any system and we can deliver it into a place where it will deliver value. So a very strong use case is going to be everything related to analytics. How do you go across your company, look at all the services that you have, all the data sources that you have, all the silos that you have, and how do you make it seamless to move that data into warehouse so that your analytics uh team can actually extract insight from it and make decisions from it. And that\u2019s really how we started. There is a ton of use case when it comes to moving data. You know, we\u2019re talking about agents these days, is like how do you get the data into agents? So that\u2019s very much what the very high-level value of Airbite is.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>2:44<\/p>\n Super helpful. And now let\u2019s go back to the beginning. You launched on GitHub. Why open source as opposed to a traditional product launch?<\/p>\n Michel Tricot:\u00a0<\/strong>2:53<\/p>\n Yeah. So when you\u2019re thinking about, let\u2019s take the analytics use case as an example. You go from like the outcome you want to drive, which is I want to be able to understand my business. The first thing you think about is okay, I will need to have dashboards, I will need to have a team, I will need to have a warehouse. And the moment you have these two, what you realize is that you also need the data, obviously.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>3:18<\/p>\n Yeah.<\/p>\n Michel Tricot:\u00a0<\/strong>3:19<\/p>\n And this is a very organic um behavior from people, which is it\u2019s not thought through so much as a strategy, but more as an enabler. So they\u2019re gonna go bit by bit thinking, oh, I need this particular silo, I need this particular silo. And it is very hard to actually think about the pain that it will be if you build it yourself, or it will be very hard also to find platforms that can support every single silos that you have. And for us, when we did open source, what we wanted is to go and talk to the team that are building all these different connectors. So when you\u2019re an engineer and you\u2019re being asked, oh, I need Stripe data to be in the warehouse, the first reflex that an engineer will have is go online, check how do I move data from Stripe, Salesforce, HubSoot or you name it, into my warehouse. And we wanted to catch these people exactly at that time. We wanted to provide them value the moment they have that little painful script that they have to write and give them something. So open source at that point is generally the best solution because I mean I\u2019m an engineer, I\u2019m a little bit lazy when it comes to if I can avoid building something, I will.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>4:42<\/p>\n Yeah, fair.<\/p>\n Michel Tricot:\u00a0<\/strong>4:42<\/p>\n And open source is generally the solution for that, and that\u2019s really why we went for like open source. The other reason is there\u2019s an infinity of places where data can be. So it is impossible for a single company to make a product that will address all the long tails of data connectors. What we need is, and what the community needs is like, in a way, all working together in a goal of like addressing all these use cases. And that\u2019s why open source for us was a solution. Like you, you know, you can think about the Linux kernel. Well, all the drivers are being built either by the community, either by by vendors, but the Linux project is not building all these drivers. They are asking the community to build those, and that\u2019s how you just get to the best uh uh product on the market.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>5:34<\/p>\n And it feels like we\u2019re seeing more and more companies open source. Do you feel that also?<\/p>\n Michel Tricot:\u00a0<\/strong>5:39<\/p>\n Yes. Um, yes, and I think it\u2019s because the technology, especially this, you know, open source is very, very present in AI, for example, because there is almost like a complete stop of the old world versus the new world. Like everything has to be reinvented. And people who are making decisions today have to catch up on a lot of context. So, what they do is actually they go talk to their team and ask them we I we need to create an agent for this particular use case. What technology should we be using? And open source generally works really well with technical profiles. And I think that\u2019s one of the reasons. There are also a lot of things around sovereignty and control that comes with open source and also future proofing because you can always update the project yourself if you want to. And to me, that\u2019s a direction that we\u2019re seeing. And having a community that backs a project just you cannot beat that velocity.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>6:42<\/p>\n Yeah, so true. So true. And okay, so you launched in 2020. When you uploaded the repo, did you know that it would take off the way that it did?<\/p>\n Michel Tricot:\u00a0<\/strong>6:53<\/p>\n No, we didn\u2019t know. We are so in the story of Hairbyte, like Airbyte started really just two months before COVID really hit the world.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>7:02<\/p>\n What a time to start. Yeah.<\/p>\n Michel Tricot:\u00a0<\/strong>7:04<\/p>\n And we had an initial product at the time, which was also related to data integration, but more geared toward marketing teams. And what happened with COVID is boom, all the marketing team got frozen, laid off, etc. etc. Because company had to figure out, okay, what does the world look like now? And you know, as a founder, you put your life into uh a company, into building a product, and you don\u2019t want to be a vitamin that I like to joke about that is not going to survive a global pandemic. So what we did is we actually went back to the drawing board. And in July, like during the period of like March to July, we were building prototypes, etc. etc. But but we\u2019re also talking a lot with the audience that we wanted to build a product for, which was data people. And all these people, they were always having a solution that they would buy, a solution that they will build, another solution that they would build, another solution that they would buy. So it was like a collection of tools everywhere just to move data. And what we\u2019ve done is just keeping in touch with all these people and keeping them in the loop of what we were building, what product. So at the time during COVID, everybody, I think a lot of people were very available on LinkedIn. Yeah. So we\u2019re very, very active on LinkedIn. So we were always trying to talk to the right people, going on a Zoom with them for like 15 minutes, 30 minutes, and then we would ask them, Do you want to be following what we\u2019re doing? And say yes. And then we created the first mailing list that we had, and every time we had updates, we would just say, Oh, this is what we\u2019re building. If you want to, we can give you a quick demo of what it looks like, and you can give us feedback. That was before we published the repo. And I think it was in November we actually put the um the repo out. And suddenly, first of all, like this initial group of people started to download the software, started to give us like real feedback, and from there it just went uh in hockey stick.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>9:11<\/p>\n Yeah, incredible. How did you feel just seeing that growth after you said it yourself when you\u2019re a founder? You put you put everything into a company.<\/p>\n Michel Tricot:\u00a0<\/strong>9:19<\/p>\n Yeah. It\u2019s uh I felt very good in a way, which is people are willing to put time into the project and the product that we\u2019re building, and yet it is super immature. And you know, we always talk about PMF in the the founder founding sphere. PMF, my definition, having seen that, is it\u2019s when people are willing to go above and beyond to make something that is not yet mature, that is not yet working, and they are willing to put the effort to make it work because it is solving such an intense problem for them that this little pain of making it work is better than the big pain of having to do it yourself. Um, and yeah, it felt good. After that, yes, I knew that the technology needed to become better, but you have to launch.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>10:08<\/p>\n Yeah, yeah, exactly. Usually, if you\u2019re at a point where you feel like it\u2019s good enough, it\u2019s too late from a launch perspective.<\/p>\n Michel Tricot:\u00a0<\/strong>10:15<\/p>\n Exactly. Like you want to get the feedback as fast as possible. You just want to build what is actually going to deliver value for your community.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>10:22<\/p>\n A quick pause to tell you about a company you need to know. ZoomInfo is the go-to-market intelligence platform built for sales, marketing, and rugoffs. By unifying data, workflows, and insights into a single system, ZoomInfo helps revenue teams find and engage the right buyers, launch go-to-marketplace faster, and drive predictable growth. With industry leading accuracy and depth of data, it gives your team the intelligence advantage to win in competitive markets. It\u2019s trusted by the fastest growing companies and has become the category leader in go-to-market intelligence. Learn more at zoominfo.com. So, how long did it take to hit PMF in your definition project market fit?<\/p>\n Michel Tricot:\u00a0<\/strong>11:01<\/p>\n I\u2019m actually splitting it because there are two paths in the life of Airbytes. There is what I call project market fit, which is we managed to create a project that was very much resonating with an audience, data engineers, data analysts, etc. And they were just taking the project and using it and contributing to it. Product market fit for me also comes when you start pulling the foundation also of uh monetization. And this is a different story because it\u2019s easy to take a product from GitHub. How you actually commercialize it, it\u2019s a different story. And to me, that\u2019s what PMF actually is. So I would say open source was project market fit.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>11:45<\/p>\n Got it. Okay. Well, take us through a little bit of the evolution then. Because you got project market fit. What kind of go-to-market decisions did you make along the way that helped you to get to product market fit from project market fit?<\/p>\n Michel Tricot:\u00a0<\/strong>12:00<\/p>\n We were not using regular channels. It was all about content. I mean, at the time content marketing was a thing, but I don\u2019t think it was as uh as popular as it has been like in 2023, 2024. But we were just always pushing articles, giving details about how the what the company is doing, what the project looks like, and just getting people to be part of our adventure. And that created trust, that created curiosity, that created a lot of awareness. You know, we published our fundraising slides, for example. So that was a way for us of like engaging the community into what we are doing. So that uh that to me is something that not a lot of people have done in the past. So it was very uh, I think very I\u2019m I\u2019m pretty proud that we\u2019ve done that. It\u2019s very, very innovative. And then, yeah, like we\u2019ve always been very strong on content, engaging with the community.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>12:57<\/p>\n So but it is time consuming content. So how do you think about that as a founder? How do you balance your schedule? When do you work it in? What\u2019s your actual cadence or was at the time for any kind of founders or operators listening that are looking to up level their content?<\/p>\n Michel Tricot:\u00a0<\/strong>13:11<\/p>\n Yeah. It is time consuming, but you know, if it\u2019s working and you feel it\u2019s gonna be working better than any other solution, you just continue and you you exploit that uh that channel as much as you can. Um after that, we yeah, we are now we\u2019re doing we continue to do a lot of content, but we are also a lot more like traditional channels like ads, uh SEO, GEO, etc. etc.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>13:39<\/p>\n So yeah. And did you write it all yourself? Did you hire a ghostwriter? Like when did you actually physically put kind of uh pen to paper, if you will, or fingers to keyboard?<\/p>\n Michel Tricot:\u00a0<\/strong>13:51<\/p>\n I would say the first year and a half, it was my co-founder and I writing. The team also was writing. So we really created that cute that internal culture of let\u2019s write something. My VP of engineering wrote an amazing article about the pain of building connectors that we keep referring to, even five years later, uh, because it really explains the pain.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>14:14<\/p>\n Yeah. Well, it\u2019s funny too. Five years later, and the pain is still the pain.<\/p>\n Michel Tricot:\u00a0<\/strong>14:18<\/p>\n The pain is still the pain, and uh yeah.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>14:20<\/p>\n Yeah. Incredible. Okay. So you lean into content early, and that helps sounds like create a bit of a tribe, and you have a very strong following of people that are passionate about the product and the space and the solution that you built. How did you think about actually taking that interest generated from your content and other means and turning it into more of a community motion?<\/p>\n Michel Tricot:\u00a0<\/strong>14:41<\/p>\n Yeah. For me, at that point, so we we also created a Slack community at the time. I think today we have about 25,000 people on it. And the way we created the community was in twofold. One is we were helping the community a lot, we are doing a lot of support because we are building the platform, and so every time someone had an issue, that was a product feedback for us. So we spent a lot of time in 2020, like end of 2020, beginning of 2020, like all of 2021, and uh we we continued after, but that was very, very intense, a year and a half, where we were always on Slack. Every single issue that was reported, we would just have something shipped the day after or like the week after. So I think that created that\u2019s that was one thing that helped uh building the community. And then what we did is we also identified a lot of champions within the community, like people that wanted to help other people. And yeah, we really engaged with them. We actually hired one of the first community managers that we that we\u2019ve had at uh at Airbite, is someone that we actually brought from the from the community that was he started to build an airflow connector, like an airflow integration, and say, Oh man, that\u2019s amazing. And we didn\u2019t ask him anything, and at some point we asked him, like, do you want to uh to do it your full-time job, like to engage with the community, write content, etc. etc. And they\u2019re like, Yeah, let\u2019s do it.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>16:12<\/p>\n So it\u2019s amazing.<\/p>\n Michel Tricot:\u00a0<\/strong>16:13<\/p>\n That to me is like the community engagement is is absolutely key. That\u2019s how you create that tribe, that\u2019s how you create that snowball effect. It\u2019s it\u2019s not something that you put on, you say, I\u2019m building community, and it\u2019s gonna happen by itself. No, it is something that has to be worked on, and you have to be intentional about what you want to do uh with the community.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>16:34<\/p>\n If you were to look back now with the benefit of hindsight, it sounds like community and content are two pillars that helped with your go-to-market motion. Yep. Is that correct? And also are there other pillars that you\u2019d say were really pivotal in your growth?<\/p>\n Michel Tricot:\u00a0<\/strong>16:49<\/p>\n Um we did a lot of events, actually, very specialized events around uh data, whether they were open source events or like Snowflake or Databricks events, it\u2019s just always getting where the people were. And that was something that worked pretty well. It allowed us to get a lot of people, new people interested, or just to engage in real life with people. Yeah um yeah. I would say, and here\u2019s really what what happened between 2020 and 2023. After that, we had we added a few other things on top of uh like how we engage with the community, etc. etc. But that to me was very very much like the three pillar of what we\u2019ve done. It\u2019s like giving a window into the company to people, giving in a window into how the the engineering team is building, giving a window into everything we\u2019re doing.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>17:49<\/p>\n And do you still operate that way?<\/p>\n Michel Tricot:\u00a0<\/strong>17:51<\/p>\n Uh a little bit less. Um but we continue to have that constant engagement with uh with people. Like, you know, when the the great thing that when you have a community like that is someone in whether it\u2019s a customer, whether it\u2019s uh it\u2019s a user, is going to ask you a question or a feature, and then you\u2019re it\u2019s gonna go into your head and say, okay, is that really useful? Or is it just for that person? And so what you do is you go in your on your community and you you just post a very simple question, like is that something that resonates with you? And in 30 minutes, you have like hundred people that are replying, yes, no, yes, but in that way. So it really accelerates how you do product discovery, how you do uh product development. So that\u2019s uh that\u2019s more like how we\u2019ve changed a few things uh along the way. It\u2019s like we\u2019re we\u2019re leveraging the the community a lot more for like what new features we should be building rather than really the the the core value proposition of the ambite.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>19:00<\/p>\n Right. It\u2019s uh a feedback loop, yeah, essentially. Yeah, great, and a very, very rapid one too.<\/p>\n Michel Tricot:\u00a0<\/strong>19:06<\/p>\n Very rapid one.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>19:07<\/p>\n So content, community events, pillars that you did incredibly well to reach the point you are now. There\u2019s always the other side of the story of you know, what were the the areas that didn\u2019t quite hit as well or almost the near-death experiences along the way that every startup goes through.<\/p>\n Michel Tricot:\u00a0<\/strong>19:22<\/p>\n Yeah, so as I said, like the beginning of the of how we are engaging with the community was very a lot of support, like helping them be successful with the product. And there was this moment where even in our in how we were working, our community became very much of a like support channel rather than like building a uh a community that was just helping each other. Um, and that to me was uh is something that we could have been more intentional at the beginning around how do we um how do we get to like community members helping each other, community members like meeting each other outside of just like rather than becoming a very much like support-oriented um uh community. And the thing is, once this habit is taken, it\u2019s very hard to shift uh into a different direction. I think we succeeded, but it took us a lot of time. We should have been more proactive thinking about okay, the community is amazing, but what is the future? Like, how do we make it more vibrant, more um yeah. How do we create a community of professionals that work in data and that are just gonna learn from each other and not just from us?<\/p>\n Sophie Buonassisi:\u00a0<\/strong>20:42<\/p>\n Yeah, it completely makes sense. It\u2019s kind of the the the tell all tales, the tell all tale story of community is a lot harder in practice, and it does require some really deep intentionality around fostering.<\/p>\n Michel Tricot:\u00a0<\/strong>20:55<\/p>\n It does, it does.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>20:56<\/p>\n Yeah, and what does that team kind of composition look like right now at Airbite?<\/p>\n Michel Tricot:\u00a0<\/strong>21:01<\/p>\n Um so we have we have a we have a DevRel person, and this this person is more um focused on the like the content strategy geared, oops, geared toward the community. And we have a community manager, meaning someone that just engages, identifies champion, uh, gives them access to um early features, etc. etc. And we also have people um in internally we call them like customer engineering, where their focus is to make sure that every product feedback around connectors is being funneled through the team to make sure that our connectors keep getting better and better and better. So this is more like for the contributors of the platform. So we really have a difference between like the users of the platform and the contributors of the platform, and we handle these two groups differently.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>21:58<\/p>\n Gotcha, gotcha. Okay. What are some other areas that you know along the journey, again reflecting back, have just been some of the most pivotal things that maybe you don\u2019t you don\u2019t see or talk about as much?<\/p>\n Michel Tricot:\u00a0<\/strong>22:11<\/p>\n Um I think that was the the realization of why are so many companies using airbite. Is it just connectors or is it something else? And connectors is a is level one, but there is a second level to it, and it took us a little by a little bit of time to figure it out, is people were also using data, uh airbite, because there was so much red tape around the data that they had internally, that having a platform that they fully control that runs within their infrastructure, it\u2019s a byproduct of open source. And we did not realize uh I would say like fast enough that that was one of the key reasons why so many teams were adopting airbytes. So, you know, when we started to to do the airbyte monetization, we said, okay, we\u2019re gonna follow the we\u2019re gonna skip the step of doing support for people that are deploying airbytes, and instead we\u2019re gonna go directly to a cloud product. And very quickly we realized, yes, cloud is getting traction, but we are not able to convert every person that is using airbite to using airbite cloud. And at that point, we just went back to the drawing board, started to talk to them, and that\u2019s when we discovered that in that case, like product market fits was not just connector. It was the fact that these pipes were under their control. And that was a big, a big thing, and I would say we we wasted a little bit of time on trying to build something fully cloud when what people needed was control and sovereignty.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>23:53<\/p>\n Got it. Okay.<\/p>\n Michel Tricot:\u00a0<\/strong>23:55<\/p>\n More like, you know, when you\u2019re searching for PMF, it\u2019s not a straight line.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>24:00<\/p>\n Never linear, never linear, no. And what are you most excited about thinking now forward?<\/p>\n Michel Tricot:\u00a0<\/strong>24:06<\/p>\n Yeah. Well, you know, every time I hear about how do I make I mean to me, like the AI wave that is happening right now is just one of the most exciting things for me and for the for the company. Like analytics is very much a core part of what we\u2019re doing, but we\u2019re getting so much pull into different types of data access. And that is something that we\u2019re today encoding into the platform and into our connectors. It\u2019s not just humans consuming data today. Yeah, it\u2019s agents that can discover what\u2019s available, discover what it looks like, and make decisions. So, yes, the technology is not yet completely mature on either side, whether it\u2019s airbite, whether it\u2019s like agency platform, etc. etc. But you can see how fast it\u2019s moving, and I think it\u2019s very energizing, especially in the infrastructure world, to see that that energy being uh being injected. So that yeah, I I I talk about it all the time.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>25:06<\/p>\n So Yeah, no, that\u2019s fantastic. And I mean you mentioned that at the very beginning around how now it\u2019s agents consuming this type of data. How does that transition in the overall industry? What does anyone need to know about what this transition actually means?<\/p>\n Michel Tricot:\u00a0<\/strong>25:22<\/p>\n Yeah. You need to you need to forget about a lot of your existing patterns. You know, I was chatting with uh with a CTO uh last week, and he told me very bluntly, I don\u2019t know, maybe he was trying to uh to be a little bit uh uh provocative here, but he said he told me, Michelle, all the technical knowledge I had stopped two years ago. I had to fully reinvent myself and reinvent my team. Uh so yes, some things are still transferable, but your default should always be thinking about how do I build in that new world? Is there a solution? No. Okay, maybe I go back and use the techniques of the of the the older world. But that\u2019s really what what I\u2019m seeing is people have to rethink how they are doing their job. Because one thing that is happening in Teams is a lot of people are using AI today to remove from their play the thing that they don\u2019t like doing. That\u2019s very easy. Like people have a very strong uh willingness to stop doing the things they hate doing. So for that, like AI is is is amazing. Like, you know, if you\u2019re an engineer, right like writing unit tests, writing integration tests, that\u2019s great, but that\u2019s just level one. The moment you actually start changing your mindset is when you\u2019re looking at the things you like doing and how can you leverage AI for those. But those are hard because the things you like doing are the things also that can bring you a lot of energy in your in your day-to-day. And those are the things that people should really be focusing on. On okay, this thing that I\u2019m doing every day, I love doing it, but can I do it using AI, using an agent? Can I ask my engineering team to build an agent to solve that particular problem? Is there an AI product that exists that can do it and removes that from my plate? And then I can focus on more things and I can become faster. But to me, it\u2019s really about reinventing um reinventing it. For data, the way you access data is very different. Yeah. Um but just having a warehouse doesn\u2019t cut it. Like you need to have like an agent does live processing, it needs to have like little pieces of data here and there. You need to provide access to the agent in a different way. So that\u2019s and that\u2019s what that\u2019s what we\u2019ve been building.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>27:54<\/p>\n Yeah, absolutely. How did that change your product roadmap overall? Did you have like this crazy moment in a way where it was like the realization that you entirely have to pivot? Or is it gradual?<\/p>\n Michel Tricot:\u00a0<\/strong>28:07<\/p>\n I I wouldn\u2019t say it\u2019s a it\u2019s a pivot because it\u2019s more like a an extension, but also sometimes we like to talk about replatforming, which is we\u2019ve we\u2019ve built the plat the platform for like a specific use case in a specific course, but there are new ones that are coming that are going to pick up massively over the next few years. And we need to be thinking about taking all the learnings that we\u2019ve had here, and how do we think about replatforming it to just have a larger breadth of use case? So that\u2019s that\u2019s more how we\u2019re thinking about it. Uh, I don\u2019t know, I would say 2024 is when we even even before like summer 2023 is is when we started to like tippito into it. But 2025 is a moment where we we went all in on that. So we still have the the analytics product, it\u2019s a it\u2019s an amazing product, but we are really building on top of that, like leveraging part of it, but also rebuilding a platform that allows agents to uh to interact with data. So it\u2019s pretty pretty cool.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>29:12<\/p>\n Super cool. And you\u2019re hitting the ground running, tons of growth, you\u2019re hiring lots of folks on the team. Like, how do you think about developing that team to take it to the next stage of growth?<\/p>\n Michel Tricot:\u00a0<\/strong>29:22<\/p>\n Oh, you have to be hammering, yeah, using AI every single day, every single or-ens that I do every Wednesday morning. It\u2019s about putting the spotlight on new uh new way of leveraging AI. And not just the layer one, which is do the thing you don\u2019t like doing. It\u2019s really about how are people building things that change their, actually change the the definition of their job. So well, if it\u2019s if it\u2019s on sales, it\u2019s gonna be around like how do they do uh like discovery of account, it\u2019s gonna be how they connect, um Like different news together, how do they connect to like past conversations that\u2019s happened on support? So it\u2019s really about like aggregating all this information in one single place and have like all the context available to them at the right time. On engineering, well, we talked enough about engineering and how agents are transforming the lives of engineers, and that\u2019s what we\u2019ve been doing at Airbytes for the yeah, for the for the past year.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>30:26<\/p>\n Yeah. And so it sounds like you disseminate this information internally. You said weekly.<\/p>\n Michel Tricot:\u00a0<\/strong>30:31<\/p>\n Weekly.<\/p>\n Sophie Buonassisi:\u00a0<\/strong>30:32<\/p>\n What does that look like? Everyone\u2019s on a team call weekly, or how are you spotlighting people?<\/p>\nDiscussed in this episode<\/h2>\n
\n
Episode highlights<\/h2>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=15<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=117<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=173<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=413<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=661<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=881<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=1162<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=1433<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=1762<\/a><\/p>\n
Watch: https:\/\/www.youtube.com\/watch?v=x3ahpXoaaVc&t=2006<\/a><\/p>\nKey Takeaways<\/strong><\/h2>\n
<\/strong>Airbyte didn\u2019t try to boil the ocean; it launched open source to solve one gnarly, universal pain: moving data from silos to value. By catching engineers \u201cat the search,\u201d they earned usage before monetization.<\/p>\n
<\/strong>Community love \u2260 revenue motion. Airbyte treated the GitHub traction as project-market fit, then built the monetization engine separately to reach true PMF.<\/p>\n
<\/strong>Publishing fundraising slides, writing deeply technical posts, and narrating the build created trust at scale. Transparency reduced perceived risk and generated consistent inbound.<\/p>\n
<\/strong>Letting Slack become a help desk capped upside. Designing for champions, peer-to-peer help, and recognition programs turned users into advocates and contributors.<\/p>\n
<\/strong>Enterprises adopted Airbyte not just for connectors but because it runs where they need it. Control, sovereignty, and security often trump a pure cloud pitch in data movement.<\/p>\n
<\/strong>Moving from OSS to hosted cloud is a different business with operational drag. Hiring too fast created noise; starting small and iterating would have preserved product velocity.<\/p>\n
<\/strong>For the first 18 months, Michel and co-founder wrote the playbook in public. Founder voice clarified positioning, attracted contributors, and set a high bar for later content ops.<\/p>\n
<\/strong>Posting lightweight polls\/questions yielded 100+ responses in minutes, compressing research cycles. Community became an always-on signal router for roadmap decisions.<\/p>\n
<\/strong>Longer cycles, more stakeholders, and ramp time are physics, not flaws. Hire earlier than feels comfortable, but in small, validated steps to avoid overextension.<\/p>\n
<\/strong>Agents are new \u201cconsumers\u201d of data, demanding low-latency access and different interfaces. Replatforming around this shift is a multi-year moat, not a feature.<\/p>\n
\nThis episode is brought to you by our sponsor: ZoomInfo<\/strong><\/h3>\n
\nFollow Michel Tricot<\/h3>\n
\n
\nRecommended books<\/h3>\n
\n
Referenced<\/h3>\n
\n
<\/a><\/li>\n
\nHost links<\/strong><\/h3>\n
\n
<\/a><\/li>\n
<\/a><\/li>\n
<\/a><\/li>\n
\nWhere to Find GTMnow<\/strong><\/h3>\n
\n
<\/a><\/li>\n
<\/a><\/li>\n
\nGTM 169 Episode Transcript<\/h2>\n