Satyen Sangani Transcript
Clint Betts
Satyen, thank you so much for coming on the show. You are the CEO and co-founder of Alation, and I've been looking at your company all morning. It's super fascinating and incredible the way you're utilizing AI in a way that makes it like an actual AI company. There's so many young startups are putting AI on their home page and really they just have a ChatGPT integration. What you're doing is really, really cool. Tell us how you became the CEO and co-founder.
Satyen Sangani
I really founded the company. Clint, thank you for having me; it's great to be here. I founded the company 12 years ago with three other co-founders. We were all people who came at the problem of understanding data from a slightly different perspective. In my own case, I grew up as an analyst and was an analyst by training, so I did academic training and work in economics; after doing that, I was a financial analyst on Wall Street, and after doing that, I built a business software for Oracle that would help companies analyze their data. Mostly banks and financial software companies. And so that was my experience. But then I found three other co-founders, one of whom was an expert on human-computer interaction and came out of Stanford, another one of whom was an expert on AI and did his PhD on fuzzy learning, and another one of whom was an expert in databases. And the two latter folks came out of Google. And so the four of us came together and broadly said, "Look, data just needs to be easier to comprehend and understand," and that's what allowed us to give birth to Alation.
Clint Betts
And what do you make of the innovation that's happened? It's not like AI is new. There was probably some version of AI going on. There have been versions for decades. But particularly 10 years ago. What do you think now about the rise of AI and how rapidly it's changing and developing? What do you think that means for your business? What do you think that means? Because AI is only as good as the data you feed it, it seems like the way you handle it and make sure that companies have access to their proprietary data are going to be the whole thing. It's really interesting.
Satyen Sangani
Yeah. But it's fraught because you would think that AI would really be able to understand what's inside of computer systems much more easily than the things that humans can understand. So the systems that store a company's general ledger, the systems that store its best customer data, things like Salesforce and NetSuite and Oracle and PeopleSoft and Siebel and Workday, all of these things are things that companies and customers use to store their most critical and strategic data. However, these new large language models are primarily trained on the text and primarily trained on unstructured files. And a lot of that technology is about predicting the next thing.
And so if you say, "Hi, what's your name?" The corpus of information would get searched, and it would say, "My name is Satyen Sangani." And so the way that computer is reasoning out is it's just predicting the next set of things given the context that it's been given. And that's really a problem of a lot of computing, which is why you see all of the spending on these Nvidia chips and these GPUs; it's really about computing and processing power. That's been the major innovation that these models have taken advantage of. But the problem is, of course, that this textual information, which is predictive and stochastic and possibly a little bit random, which is why you see all these hallucinations, doesn't necessarily speak well to your business information systems, which rely on precision and which rely upon having a single definitive answer. If somebody says, what are your profits, it doesn't need to be like, well, it might be possibly 3.72443. You don't want it to guess. You want it to know. And so where Alation sits is in the middle between these two things. We enable companies to figure out how to get these really unstructured models, these generative models, to talk to business's most critical systems and really get something that's comprehensible and reliable out.
Clint Betts
What do you think? How rapidly is AI changing to the point where? How much time are you spending studying the changes and figuring that out versus building your product and focusing on your customers? Is it really going as rapidly as people say, or do you have the time and ability to really focus on what you are doing, your product, and your customers?
Satyen Sangani
I think that's the only thing you frankly can do. You obviously have to understand and leverage the biggest, best, and most interesting tools in the market. And Certainly, that's coming from OpenAI and Google Gemini to a lesser extent, and folks like Anthropic and then even Meta are obviously doing a lot of interesting work with open source. Those companies are pouring literal billions of dollars into building and training some of the best models. So unless you're massively well capitalized, unlikely to be at that bleeding edge of the technology. So then the question really is, well, what can you learn and leverage from the latest models? What's interesting is that those models themselves are moving so quickly, but even the older versions of the models, in many cases, are ones that we haven't quite figured out how to take advantage of fully because the technology is so nascent and so new. But the best thing that you can do is really figure out, okay, what problem my customers want to solve. What are we trying to do? How can the model help? What are the use cases that we can use to apply this technology? And I think that's true for ourselves, and frankly, it's true for most people who are now consuming and experiencing and interacting with these technologies.
Clint Betts
What do you think about AI governance, given the importance and sensitivity of this data?
Satyen Sangani
Yeah. So, one of the big challenges that all of these enterprises have is that they don't actually even know what's being used and where these AIs exist inside of their companies. Who's using ChatGPT? Are they using it for customer interactions? Are they using it for programming? Who's using these other models? Are they using it for internal productivity use cases, or is it being used for external applications? Are we buying third-party applications that contain AI? And what aspects of AI exist? Now, one is just an existing question: Where are we using this stuff? And the reason you want to know where we're using this stuff or whether we're using this stuff is because you are now letting a machine intermediate what previously a human might've done or accelerating work that a human might've themselves had to sweat. And so there's a real risk that the machine is getting it wrong. And so you need to know exactly where these things are being used so you can figure out whether or not there's a massive risk or significant risk or small risk in making sure that your business is not adversely affected by these AIs.
Obviously, an agent or an LLM can make some very innocuous, easy decisions and suggestions, but it's also the case that they can make really bad recommendations that could be damaging to your reputation as a company or could be financially damaging. And so knowing one, where the existence of these models is, two, the data that are feeding these models and the versions of the models that you're using, and three, who are using these models and how they're being used in a company is just table stakes. Now, I would argue that most companies don't even know those things. Today, what people need is essentially a form of governance. People talk historically a lot about this thing called data governance, which is really having awareness and the ability to apply policies, regulations, and rules to the use of data within a company. Very similarly, you've got to do the exact same thing with AI.
And there are starting to become regulations. There was a recent California bill that was just actually not passed but came close to passing. I think it was SB 1407. And there's also, of course, the EU/AI Act, and all of these acts are basically being put into place because people need to be able to identify what AI exists. In some cases, some of these regulations say you have to put a kill switch on this stuff. And so if you're going to be a company that's going to use this stuff, which you have to because you have to use it to stay competitive, then you've got to basically know what you've got and how you're going to be able to control and manage it.
Clint Betts
Can you give us a typical use case for somebody, a client, or a customer of Alation and how they use your product?
Satyen Sangani
Yeah. So often, we are used by technical teams, data teams, and business teams. And so people who basically use data. Interestingly, there are different ownerships or lenses of data. If you think about what data is, it is just an electronic representation of the world of my business. A customer contact list is a slice of my business. My general ledger is a slice of my business. All of the visitors to my website in the last 48 hours are a slice of my business. You can piece all of these slices together to reproduce that elephant. There's that old tale of the blind men and the elephant. And so what people are trying to do in these very complicated businesses that mostly are interacting with digital customers or complicated customers or customers all over the world, they're just trying to reproduce what really happened.
Now, the problem in reproducing is what really happened is that you don't necessarily know where the data came from. You don't necessarily know the assumptions that the programmers had when they were producing this information. You don't necessarily know what the term definition means. So somebody might say curbal_c; well, what does that actually mean? What does that mean, and how do I actually understand how to use that particular thing, and how does that differ from curbal_c? Maybe one is the foreign currency balance, and the other one is the domestic currency balance, but you wouldn't necessarily know that if you were just coming to some random table in a random database.
So what Alation does is it helps people really collaborate around this information. It helps people understand what those things mean. We are an incredibly modern, sophisticated platform in that we bring all of this AI and all of this cloud computing to bear to basically translate data speak from technologists into human speak for people who are trying to motivate and run the business forward. And so we're really a translation tool as it were, but also we are a collaborative knowledge base because these companies grow up over years and years and years with all of these different systems, thousands of systems, tens of thousands of tables, and it's really difficult to reproduce what happened and we help companies reproduce what happened and understand the data that helps them understand their own businesses.
Clint Betts
What do you read throughout the week, and what reading recommendations do you have for our audience?
Satyen Sangani
I have a diverse diet. I do listen to podcasts, so there are a couple of different ones that I really love. One is actually The Rest is History, which is a podcast about various historical events, and they both dive in deep into various people and various events in history, which has been awesome. I do get news from a couple of different places. The Economist is one, the New York Times is another. It's hard these days, though, to figure out what news diet you're going to get because sometimes stuff is really biased, which is true for data and true for news. And then I do a lot of what I'll call grazing. I'll just find different stuff on the web and different articles about what's current. I am a big book reader, so the two recent books that I read, one was by a guest that I had in my podcast, a guy named Chris Wiggins, who's a Columbia professor, called How Data Happened, which was about the birth of data. And then another one, I just went to Chicago with my son and read a book about The Devil in the White City, which was a really great history novel. I would say I'm pretty eclectic in my taste, but hopefully, I try to keep it interesting. I like to read stuff that expands how I think about things just as much as I do about my regular day-to-day data stuff.
Clint Betts
Oh yeah, I love that. What does a typical day look like for you then as CEO of this company?
Satyen Sangani
I try to wake up every morning, and the first thing that I've done ... And this is actually a new habit, like 60 days old. Is to just do something for my body every single morning. So whether it's like a yoga session or a workout. On days that I'm super busy, the permission that I've given myself is that I can be just 10 minutes long, but I do something just to get my body moving. I also then have a ritual of making myself a cup of chai, which is pretty meditative for me because it allows me to really focus on how I do it, which sounds silly. But literally, the act of just putting in the spices and doing it in a certain sequence just sets me up for the morning.
I will then check my email, so I try not to check my email first thing. That's been a pretty good hack that I've acquired over the last four months, which has changed my early outset for the day. And what's interesting is a lot of these habits are new. I would say that the anti-pattern had historically been waking up in the morning, looking at my phone first thing, starting answering emails first thing, and not doing that has been a real, just incredible unlock in terms of just my own mental thinking and framework. Then, dive into my day. The day can start with a variety of internal and external meetings, generally meeting with at least a customer, a partner, or a couple once every single day. Typically, we'll have team and organizational meetings. I have some regular meetings, which are one-on-ones, but I mostly do a lot of meetings in a group setting. As a CEO, I really want my entire team to hear what each other is thinking, as well as what I'm thinking. Then, towards the end of the day, I try to schedule some of my work. But I would say one day a week; I'll try to block out space so that I can really just do my own reading and writing and not have to be in meetings all day.
Clint Betts
Yeah. As you think about culture and building a culture and building values, how do you think about that? How have you built that inside of Alation?
Satyen Sangani
I think culture is fundamentally defined as the set of habits and values of a people. So culture isn't being ... Well, it could be. It could be super being super enthusiastic and expecting everybody in the company to be super enthusiastic at all times. And then you're like, everybody's rah, rah, and we're all enthusiastic. We love this place. And so it could be that. That's not our culture. I think a lot of what you're trying to figure out is what you reward and what you recognize. And in some cases, what do you criticize or punish or at least correct? And so a lot of what I think about is when people follow the right process, when people are being brave when people are supporting each other when people are going above and beyond and doing things that are really hard work. And those are the things that you want to try to celebrate.
And I think what we see inside of Alation is that there are a whole bunch of employees who do the same thing for each other. And so a lot of what people talk about when we do employee sat surveys is, what do you love about Alation? Do you like the all-hands? Do you like the food? Do you like the office? Whatever it is that people can possibly like. The number one thing through the years has always been, well, I really like my colleagues, and I really like the people. And I think it's because people have this shared value of really wanting to make progress and be ambitious, but also that they're somewhat supportive of each other. And I think that balance is what I really think about trying to get right.
Clint Betts
How have you thought about working from home versus in-office versus hybrid at Alation?
Satyen Sangani
Are fumbling through it. But I think making a ton of progress. Just starting this last month, we have brought people back two days a week. It's pretty likely that we'll move to three days a week starting at the beginning of the year. We've yet to make that final decision. But what we found is that the energy in the office is actually pretty incredible. I was actually a little bit worried. I was worried that people would see it as a drag on their time, that they wouldn't, of course, want to commute, that it's a massive change of habits. Certainly, if you're a parent with kids, that makes it so much harder because you've got pickups and drop-offs, which you're working from home could be a little bit easier if you're at home. But people need to be connected to each other and the work that they're doing. And there is so much communication that just doesn't happen. Even in this setting where we're on Zoom, you just don't get the connectivity that you can get with the bandwidth that you do if you're in person. You don't see whether the person is slightly annoyed, whether they're looking necessarily at their screen because they're looking at their email or they're looking directly at you. Those moments actually create so much more connection and trust, and it's been really incredible to see.
I used to be super biased towards making sure everybody came into the office every day. And then, of course, with the pandemic, we all had to adapt, and we went to a remote first company. I will say that I'm probably full circle on this, and I'm of the view that people just have to be together. And especially when you're trying to innovate. There are some moments and some jobs where you're really trying to optimize for individual task productivity, and that person doesn't have to collaborate. But that's so rare in the modern workforce in terms of knowledge workers and what we have to do. Most of the time, what you're doing has to depend on somebody else, or it does depend on somebody else. So yeah, I'm a huge proponent of building that connection.
Clint Betts
How have you thought about raising venture capital? If I understand it right, you raised in 2022 $123 million series E led by Thoma Bravo, an incredible fund, actually. You're valued at 1.7 at least at that time. 1.7 billion. How have you thought about that and how are you thinking about the public markets right now? You get to series E, you're starting to look at that, I'm sure.
Satyen Sangani
Yeah. I remember when I first started raising my seed round, and I remember going to meet with investors, and I was so nervous. I would be on Sand Hill Road, and I would park in the parking lot of these venture capitalists, and you'd just be super geared up and, in some ways, anxious and nervous about meeting with these people. It would be life and death in some sense because you obviously needed to pay your salary. And I'm privileged enough to have gotten past that first milestone and four or five fundraisers since. And it's funny because now, so many years later, having been on the other side of that, fundraising is a means, not an end. We all celebrate the idea of raising money. It's obviously a thing that makes the news headlines. People talk about it all the time. But the reality is that you're using these funds in order to produce value. And the act of actually getting valued doesn't itself produce any value. In fact, the day after a fundraise as an entrepreneur, you strictly own less of the same exact thing you owned, plus a little bit of cash or a lot of cash, depending on how much you raised.
So I've really increasingly felt like one: we over-celebrate fundraising. But two, I think that it's something that you want to really use as a conversation. You obviously need to sell. As an entrepreneur, you've got to sell your ideas. You've got to distort this future version of reality, as Steve Jobs talked about, in order to be able to explain to people why the world is going to move in the way that you're going to build it. But for me, it's always been a discussion of the problem statements. What are the problems that we are trying to solve? Why are we the best people to solve those problems? Why have we done the best thought? Why our product ideas are the best. And if you can make it a conversation with your investors, then you're going to attract the right type of investor versus somebody who you're selling just a bill of goods and who, at the first sign of something going wrong, will ditch you.
So, I think the best investors and the best entrepreneurs are partnered together to figure out how to make something bigger. And if you can find that investor and think about it as a conversation, you're going to have a lot more credibility than just trying to sell somebody your idea and not listening to their feedback. But much more critically, not listening to the fact that both of you are trying to create a shared outcome together.
Clint Betts
What have you learned about how your leadership style needs to change as the company grows? The early stage is much different from being a leader of that company, as compared to where you are now. How have you thought about that and the progression you need to make as a leader?
Satyen Sangani
Yeah. All of these new articles discuss this age-old issue. So Paul Graham, who started YC, had this founder mode manager mode comment, and then people are like, "Oh no, it's not founder mode and manager mode; it's owner mode." All of these things speak to the idea of how much you delegate and let other people run with versus how much you yourself get into the details. And I think early on, I was just a strict micromanager. I'd walk around, and I'd love Google Sheets, and they're running jokes inside of the company around how I'd have task lists and to-do lists. And it was definitely the case that I went way too long with the spreadsheet and the to-do list and running team meetings with to-do lists. And that was definitely a mistake. Especially if you want great people. Great people do not want to necessarily be told what to do. They want to be given a problem that's capable of solving and then run with that problem so that they can figure out the solution to it.
The other side, of course, is going to full-on delegation, and often, you hire executives and individuals who just might be way, way, way abstracted from the problem. Might not get into the weeds. It might just be like, "Look, my entire job is to hire people, and I don't really necessarily get involved." And that doesn't quite work, either. And as a CEO, you can fall into that mode. And the real thing that I've found is that at any given moment, there are one or two or three initiatives that you have to drive. There are things that just have to get done. They're the most critical things in the business, and you have to pick your own state and go arbitrarily detailed on those things. That's the Elon Musk. I am sleeping on the factory floor at the moment. But you've got to pick those spots and to drive the business forward, you've got to make sure you get those things right.
And then there's everything else where you really want to trust and delegate to your team and do the rational things of QBRs and business reviews and making sure that everybody's coordinated and really rely upon management best practices by doing one-on-ones and team meetings and OKRs and metrics. That abstraction is really helpful, but you can't totally abstract from every single thing out there and not have any project ownership. Because as a founder, the one thing you have as a CEO, the one thing you have is the ability to drive change forward. And so that's been a big learning for me, which is to be really judicious about where to pick the spots. And I struggle with it all the time because you're constantly slippery-sloping either into too much management or too much delegation. Or, on the flip side, in some cases, doing too many detailed reviews about things that really don't matter and that your team can really figure out.
Clint Betts
The reverse is that a lot of CEOs now are being asked to comment on things that are happening outside their company. Basically, cultural or societal issues or things like that, which maybe wasn't the case a decade ago. How have you handled that? How have you thought about your role in society now when it seems like business leaders are being asked to comment on things that are outside of their company?
Satyen Sangani
It's really challenging. And like many of these things, whether it's back to office or culture, we all are learning and trying to figure our way out. And I don't think there's a single clear answer. On the one hand, there are people who think that they're going to opine on everything. Every political event is something that they've got to necessarily comment on or at least the ones that they feel strongly about, and they're going to get out there, and they're going to talk about these things.
On the other hand, there are people who say, "Look, I'm not going to comment on a single thing." I would say that probably the year is more towards the latter, and I'm more inclined not to want to share my political viewpoints than I would be to try to share everything. And for the sick reason that, frankly, I'm not a politician. I don't have awareness or depth of knowledge of the most recent issues in all cases. Even though I might read a lot, I probably don't read as much as these people. I'm certainly not elected. I'm not elected by my customers, I'm not elected by my people. We are a for-profit corporation, and the people that elect me are my board. My board pays me to make them more value for their shares. And that's, I think, largely what my employees care about as well, in addition to building a great culture. But I think in building a great culture, there are certain moments where you do have to be better than silent.
One example was the George Floyd case. We put out a blog post that I co-authored with my CPO at the time, my chief people officer. And there were a lot of our African-American employees who were really hurting. They were just really struggling with who they were, what their identity was at the moment, and how they were being received and perceived. And that was a really important moment for them. And so I think you do have to be empathetic in moments to your people, but I think it's really to your people. Most of my customers don't care about my political opinions, and they shouldn't. I think it's a balancing act, and you've got to really pick, and there's no right rules. But you've got to think about your constituencies, what your business does, and where it lives, and then I think you should pick your spots, but be pretty careful about doing it.
Clint Betts
We're getting to the end of the year here, end of 2024. And obviously we have an election coming up. I don't know when this will air so maybe it would've already happened. How are you thinking about 2025 from a market perspective, from a business perspective? 2024 did seem like it was a little bit hard for tech companies. Do you expect that to continue or how are you thinking about and planning for 2025?
Satyen Sangani
We are expecting a little bit of the loosening of the markets, but frankly, not a massive change from the current state. Certainly, I think that whether there's a Trump presidency or a Harris presidency will impact tech M&A more than anything else. And it's not clear where Kamala Harris would go as far as the FTC is concerned. However, to the extent that antitrust regulation and tech mergers are less prohibited, it seems like the Trump administration would be much more active. And I think that would make for a really exciting environment because if there's more or less consolidation, that would change the business dynamics pretty robustly. So I think you could see a situation where there are dynamics in the business changing. I think with interest rates falling, you might also see the threshold for going public decline a little bit. You might see a little bit more spending.
A lot of spending has been allocated to AI. Unclear whether that will arrest or whether that'll simply accelerate. I can see both sides to that story. But we're not planning for it to be, I would say, some reversal, if you will, to the 2022 days or the 2021 days frankly, where it was zero interest rates, and SaaS companies were the best business model in the world, and everybody was simply spending on SaaS and not even questioning that spend. I do think that spending is going to be much more judiciously measured, and that's, I think, a really healthy thing. Frankly, there's a lot of software in the world, and it's really hard to make all of it work. And if you're a CIO, you've got to figure out how to make all of this stuff function in your environment and actually get value. And just because a computer program exists, it doesn't necessarily mean that it's going to make your life better or your customer's outcomes better. And so you've got to be really thoughtful about implementing and delivering this stuff, and I think that's how it should be, frankly.
Clint Betts
Finally, we end every interview with the same question, and that is at ceo.com we believe the chances one gives is just as important as the chances one takes. When you hear that, who gave you a chance to get you to where you are today?
Satyen Sangani
Ooh, so many people. Do I pick one?
Clint Betts
Whatever you want. Yeah. Go ahead.
Satyen Sangani
Three people come to mind. One is my high school history professor. A guy named Chuck White. He just inspired me with American History, had a huge impact on my going to college where I did, and was just somebody who really deeply believed in me. Oh man, the two women in my life who are just incredible are my mother and my wife. Both of them have done all of the work to build our family and to believe in me. And my wife was our first venture capitalist. She's the reason why I even started the company in the first place. So, certainly, the two of them get far less recognition than I do.
The journey of being an entrepreneur has been incredible, but as I go back to thinking about fundraising, the person who's walked every step with me in the journey has been a guy named Greg Sands, who's our lead investor from Costanoa Venture Capital. Greg is the first call when there's a problem, and he's somebody who, no matter how bad things are, will always bring perspective and will believe in you even in moments when you're a little bit unconfident about yourself. And so, gosh, but there's so many people. The thing about being an entrepreneur is that you just walk the first step, and then if you have a good enough idea, there are a million people who will walk it with you. And that's every one of your employees, and that's every one of your customers. And so, in some sense, the entire process of entrepreneurship is just an exercise of common belief. And if you've done it right, those people will pick you up when you're not able to go a little bit faster and farther, and hopefully, you do that a little bit as well.
Clint Betts
Thank you so much for coming on. Seriously it means a lot that you take the time and best of luck with everything and best of luck with 2025. I do think you're right about the loosening up a little bit, but it can't be all rosy like it was in 2021, '22. That was crazy.
Satyen Sangani
Yeah, absolutely. Clint, thank you for having me. It was a lot of fun.
Clint Betts
Yeah. Thank you.
Edited for readability.