Tristan Handy, dbt Labs | 2025 Tech Innovation CUBEd Awards
>> Hello, welcome to theCUBE here in Palo Alto, California. I'm John Furrier, your host of theCUBE, and we are here covering theCUBE's Tech Innovation CUBED Awards where the winners were CUBEd with an inauguration event that we've been doing. And since the program that we identify the innovators and the winners get the CUBED award. The competition was fierce.
We got a lot of great entries, a lot of exceptional use cases showcasing the innovations across the industry. And today we're joined by Tristan Handy, the founder and CEO of dbt Labs, to discuss their winning entry driving the industry forward. Tristan, congratulations for winning the awards.
Obviously data is a hot category, what you guys are doing, we've been following since the inception, the whole evolution of the data platforms, so congratulations to you and your team for winning the awards. And- >> Thank you so much. Great to be here. - ... let's get into it. So the top honors in this area is interesting.
Now what I love about your category is it's the crossover between platforms and AI. That's my opinion. We could debate that or talk about it, but as you look at surge of AI and the data models that are coming, a lot of the developers are looking at the integration of data and data models that's on top of the existing data sets that are out there, because AI wins with data and data availability and synthesizing that data, making that available and secure. Data lakes are popular. People are throwing data into a data lake, and so you're seeing a lot of activity, but now you've got to stitch it together and then the hype of agents is another opportunity. So you guys are in the middle of it.
Can you share the key innovations that was central to your award-winning entry? >> Oh, geez. You already started on AI. I want to talk about AI, but let me go back in time a little bit to answer your actual question here. I would say that the past 10 years, what are we in 2025? Past, let's call it 13 years, have been about the rise of the cloud in data. And that starts with the launch of Amazon Redshift and then soon thereafter, Snowflake, BigQuery come out and then there's now a big field of folks. And all the way back in 2015, I was looking at this using these products and had this, I don't know if I can say holy shit, but holy shit moment all the way back then.
And the cloud products that were available in this category were just so far and away better than anything I'd used before and made me want to jump in and help clients adopt this technology. And there was an obvious gaping hole in the ecosystem of data transformation. There was no cloud-based way to do it. And so I think if you want to talk about innovation, innovation is not the nuts and bolts thing that most companies do every single day, and that includes us. Most of the things that we need to do on day-in-day- out basis are take the core innovation and extend it to what customers need, how to make it useful in their context. But I think the core innovation for us was in the 2016, '17 window when we were fundamentally saying what we were going to build, and that was a cross-cloud layer that used SQL to do data transformations and allowed data engineers and what we call analytics engineers to bring the best practices of software engineers like CI/CD, Git, all of these kinds of practices into data, which previously is not an area that had any idea of software engineering best practices previously.
>> Yeah, I mean, you bring up data engineering, again, we're of like minds, we've been riffing on theCUBE, " Data engineering is like platform engineering, it's like SREs for servers. " You started to saw that horizontal scale. Talk about the role of data engineering right now because I think that's a key point that we're starting to see a lot of that same pattern emerge as data becomes the lifeblood of the apps now because now with generative AI, it doesn't exist without data.
So if you're not set up, again, cross-cloud, we called it super cloud because we saw that abstraction opportunity, this is a holy grail moment, it's a holy shit moment and a holy grail opportunity in the sense of this could be the moment where if you get it right, good things happen. And it's not just software, it's engineering, right? So the engineering is a big part of it. What is that data engineering formula? If you could share your thoughts on this, because I think that's a key point. >> So I think that most jobs in software go through this productivity curve, and if you look at app development, you could go back to the mainframe era in, whatever, '60s, '70s and say it was obvious that we needed business applications for a million different things and every business in the world needed a bunch of business applications. But the underlying software ecosystem just didn't make that possible.
Mainframes were too expensive, writing new programs in mainframes was too effortful. So you have to get to this point where actually the cost- benefit equation makes sense so that you can actually deploy people against these problems and build stuff. I think that data engineering has gone through that productivity curve as well.
Data engineering is not new. If you go back and look at the beginning of modern data engineering, I would personally put that at the dawn of internet search where everybody indexed the entire web and put it into a single place and that's a massive data engineering challenge that is like what we do today, but that's also, it turns out to be the cornerstone of a multi-trillion dollar company. And so yeah, you're willing to go through the hard stuff to get there. The problem is that most enterprise applications don't present trillion dollar opportunities and so you actually need better productivity tooling. So what's happened over the last, I think, decade is that tooling, frameworks, the entire software ecosystem around data engineering has gotten better.
And so what we've seen is a lot more data get moved and a lot more data products get built on top of that and that feeds to the current moment where we are right now with AI. We still have further to go there, but I think we're in a better place than we have been to actually feed structured data into AI systems. >> What are you guys looking at now in terms of the opportunity and problems you're solving? What is the state-of-the-art deployments? What are the challenges and opportunities you guys see? Because we're now at a point where there's general consensus about how data is being used. You mentioned some of the toolings getting better, platform engineering, cloud native is booming. We're now on this next wave, call it AI, whatever you want to call it, but what are you guys doing now? What's the core state-of-the-art for managing data across the silos, across clouds? What are some of the customers looking at doing? Are they like in early innings, fourth grade? Are they more savvy? Are they looking for integration with existing stuff? What are some of the things you guys are working on? Share your customer mindset, deployments, operational challenges.
What are some of the things? >> Totally. I talked about the last 13 years. I really think that starting whether you want to count it last year or this year, the next era in data is not going to be defined by purely the move to cloud. I think that it is going to come from really two trends. One is standards and the other one is AI. On the standards front, we could talk about this for a long time, but the thing that happened last year that was what I think of as the shot heard around the world was two back-to-back conferences, the Snowflake and then the Databricks conference, both in San Francisco in June. The CEOs of both companies got up on stage and they said, "We are committed to Iceberg as a standard way to store data in the cloud data lake.
" This presents an unbelievable opportunity for companies throughout the space and for customers because I think if you go back for basically forever since the rise of the relational database and probably before, the source of profits in the data ecosystem has been lock-in around the data storage. You put all your data in a certain format and a certain engine and then they charge you for access to that. And for the first time, that changes. And so what we are very, very quickly seeing is that went to the very top of CDO's lists overnight. We want to understand our strategy with respect to Iceberg and OpenTable standards, and it gives us the opportunity to bring together organizations investments across data platforms and have single views of data as it flows through that organization. And I think that's- >> So open standards are critical for you.
You see that as an absolute game changer. Or not. >> If you don't have the ability to access the same data sets from multiple different engines, then you're always going to be fighting with the laws of physics. And data actually has a physical component. It's stored on drives, on SSDs, and there's atoms involved there.
And moving the state of data around, it actually costs money. And so open standards actually enables the cross-cloud world that we've all foreseen would happen, but we didn't ever quite get there until now. >> One of the things about the awards that was interesting was we prompted it as, what were some of the innovations that you guys are doing? You guys obviously won the category.
Looking back at the journey, you mentioned some of the history, what you're working on today. Was there a turning point or decision that you guys made that got you this recognition today? That's obviously just a recognition award, but you guys, you're in business to help customers. Was there a moment in time, was it 2016? Was it post-pandemic? Was there a moment where you're like, okay, there's a turning point here, we see it go all in? Can you share your thoughts around that journey or key moments? >> Yeah, I mean any hockey stick curve is built of many, many small decisions.
The decisions that were made in the first six months of the product being used by the community certainly were very impactful. I would say that if you want to choose one single moment in time, it would probably be in the second half of 2019 when we first really considered changing our business model, and we didn't get into this, but we started off as a consultancy and dbt was a side project and we can get into that if it's interesting, but dbt caught fire in a way that we had not anticipated and everyone was telling us it was a bad idea, but more and more- >> Usually <inaudible> sign, it's a good idea when everyone thinks it's a bad idea. >> So yeah, slowly more and more humans and data got exposed to dbt and started using it and we had this anonymized instrumentation data where we could actually see the usage climb and climb and climb. And by mid to late 2019, we had about a thousand companies using the product, paying us basically $0, and supported by three or four full-time software engineers.
And that equation just didn't work. We needed a way to invest more and that moment in time where we had watched the graph go up and to the right for three years at that point. And so we said we feel enough confidence in where this thing is going that we're going to move to the venture funded model and make this thing grow. >> Well, I would love to do a deep dive on that. I love that story. I think that's a
great case study and template. I mean all successful companies start by sometimes by accident or side project or sometimes you get into the market and you navigate through it and just, you get product market fit. So I'd love to drill down on that.
On the award side, that's just proof points that you guys are doing well with customers, so it's a great innovation there. I want to ask you about the culture because I'm personally fascinated by people who do well. Every company has a cultural thing.
Moore's Law was Intel, operational speed, every company has a secret sauce. What would you say is your distinct aspect of your culture, your mission that's distinct to you guys that has enabled you guys to push the boundary? What would be that secret sauce or culture DNA? >> We have a bunch of unusual values that came from the beginning of this journey and how originally it was just me hanging out a shingle and doing some consulting work. The one that I think is maybe the most unusual is our value called We Contribute to the Knowledge Loop.
And the knowledge loop is a term that was introduced by, oh shoot, I'm going to forget his name, a venture capitalist at Union Square Ventures, and he talked about how really building successive layers of human knowledge is really what separates us from the very first humans. And we've baked that into our DNA. It's why we default towards open source and towards communities.
One of the frustrating things that you experience when you work at a failed startup, and I've worked at one or two of those in my life, is that it's this intensely frustrating experience to build a thing that ends up not being commercially successful. And then when it eventually gets acquired or shut down or whatever, all of the work that you did just goes away. It just, it's like it never even happened. But when you open source code, when you build communities, contribute in public, that stuff never goes anywhere. Every single thing you do can benefit humanity forever.
>> Yeah, and I did an interview 15 years when we started theCUBE, Doug Cutting at Cloudera. Same philosophy. Open source is the only way to go. Proprietary, it dies with the company, all that work, all that knowledge, there's no leverage. So totally on the same page there. Great culture. To wrap up, I want to just get your quick use case. Walk me through an example of a customer you're proud of or that illustrates some of the things that you are doing that recognizes some of the award points.
Is it value, revenue, is it efficiency? Walk me through a specific example of how your customers are using you guys to solve their problem that wasn't possible before. >> Yeah. One of my favorite customer stories with dbts are our customer, Siemens, obviously huge manufacturing giant out of Europe. They have hundreds of thousands of employees and thousands of data engineers, tens of thousands of data workers, knowledge workers of different forms.
And you can imagine in that kind of environment, chaos could be the default state. And the thing that I have been so proud to build alongside the team, the central data team at Siemens, is a world in which dbt is a part of the core data platform and it forms the fabric for how these distributed teams and even distributed companies in the Siemens portfolio interact with one another. So you join the Siemens data engineering team and dbt is a part of your tool set. And when you look at a chart of how data flows inside the business, it is almost overwhelming at the complexity there. But that's how business gets done and we're super proud to help facilitate it in a non-chaotic way.
>> Your vision and vibe on open source, knowledge loops, when you look at data, it's also a connective tissue between people and truth. >> Yes. - And so, one of the big themes coming into AI just to wrap up is as data becomes part of this connective tissue or network on top of pre-existing stuff, databases, whatnot, it connects people.
So Siemens groups can get better fidelity around the data. And truth is big right now. Trust and truth are really big principles right now in AI, given where the gaps are.
Just what's your vision on as these evolutionary times happens, abstraction on top of other things, we've seen this in tech, IP addresses and you got domain names, so you always have these, the abstracting away the complexity layers of innovation, trust and truth connect people, entities. Data is a binding and also a distribution dynamic. There's two things going on with data. What's your vision on this? Because it builds on knowledge. Knowledge is open. If it's open and connected with context, magic happens.
What's your vision around that? Do you agree? What's your reaction? >> I had never thought of the comparison to domain names and ICANN. Domain names wouldn't work without ICANN. And so you need those arbiters of truth in the process to make sure that the totally decentralized system actually works. You're totally right. The thing that AI systems are hungry for is context. So the way BI used to interact with structured data was that the database held all the data and just lightweight context about the data, like data types and this kind of stuff.
But all of the actual context was held in the head of data analysts and they went to the data, they brought that context to the data and helped the rest of the organization understand what was going on in there. That doesn't work in AI. If what we want to do with AI is give access to that data, to the much broader set of knowledge workers that doesn't spend every single hour of their day thinking about data schemas, then you have to take all the context and you have to load it into the heads of the models, the context windows of the models.
And so I think that the real question is how do we create enough context and how do we connect that structured data context into the models? And I think there's good answers there. I'm following pretty closely. OpenAI just recently launched a new API to provide context. Anthropic has a protocol called MCP, model context protocol.
I think these are going to be significant unlocks. And governance, who can do what, has to be a part of that. >> Tristan, congratulations. Love what you guys do. Love your vision. I think being in the data business doesn't mean you have to be a database person. You just have to be just thinking holistically around the system, the engineering of it, how the coding might work, and in general environment, how applications will interface.
Again, really, really on the front lines of where we're going. Again, more models are coming. Soon, it'll be just one model, just data in the future.
So congratulations and thanks for participating and congratulations on the award you guys got. So you've been- >> Thanks so much. It's an honor. >> Great to have you on. Thanks for coming on. Appreciate you. >> Thanks. - Okay.
Tristan Handy, founder and CEO of dbt Labs, really innovating around data and they're looking at the architecture of what the platform will need to be and what it can become in the future. We're going to have generative data. We're going to have operations changing.
The impact of data to our world is fundamental to society, to human rights, to everything going on. And theCUBE has got it right here covered. Thanks for watching.
2025-04-02 01:41
