The Exchanges

Q Tell us how you make money. What's your business model?

A Yeah. So we're a subscription business. So we actually call it data engineering as a service is kind of the current offering. Um, so we're building a platform out, but while we're building the platform out, we're basically doing this work on behalf of customers. So it's sort of a cross between a services company and a software company at the moment. Uh, and so we, we basically offer access to our team. And our technology for one flat price, you know, uh, six grand a month if you want it to run on in our shared cloud, or 10 grand a month if you want us to put up a private instance of the platform so that your data, so that your work is all separated from other, uh, companies.

AI assessment note: “we're a subscription business... six grand a month... or 10 grand a month”

Q know, what prompts someone like you, I mean, everything we know publicly about a genre, I mean, when you left in 22, this thing is, it's doing great, you know, it's in the dev tech space, you know, it's, it's, it's grown like a weed. Wayne, why, why decide to sort of leave? Were you maybe just bored or something else? Can you share your story and transition to Tembo?

A Yeah, I mean, I tell a lot of people this, especially in the entrepreneurial ecosystem. I was, I was, I was offered a chance to sell my shares in secondary at our Series B, Series C rounds. You know, that was, this was the Zerp era, and I said yes. You know, I was basically, the way I saw it was, you know, I'm a founder, you know, sitting at a, at a poker table with a big giant pile of chips, and I have no, no money in my bank account, you know, and someone says, would you like to leave the table? It's like, Uh, yeah, let's, let's make this, um, equity real. Uh, so I had a rare chance of making it happen without, without the company being acquired. Um, you know, we just had a lot of interest in our, in our rounds and, you know, I'd been doing that, that company since 2015. So, you know, seven years in, um, I also had hired a CEO, you know, we hired a CTO, we hired a bunch of leadership. And so, you know, I just wasn't, I'm, I'm an early stage person. Um, I'm a from zero to blank, uh, person, you know, not, not what they're doing now. It's not my, not my talent.

AI assessment note: “I was offered a chance to sell my shares in secondary at our Series B”

Q So, uh, okay, good. So give us more of the, kind of the history. What year did you launch the company in?

A We launched the company in May of 2015. Uh, we were actually at a, the collision conference in Los Angeles, or I'm sorry, in Las Vegas, and, uh, decided to pivot it after the first day of the conference. Uh, well, we had a booth, and people were coming up and talking to us, and we didn't like our company. It was the first time we really got out of the, out of the building, and talked to a lot of people at once about what we were doing, and realized, uh, it was kind of funny. I mean, we just realized, like, we didn't quite have the passion for the problem, um, And decided to step down to a different problem. So we were building an analytics company similar to what Mixpanel and, and, you know, some of those, uh, sort of tools do, and realized, like, our biggest problem was, um, getting companies to send us the data to, to run these analysis, um, and, uh, And just recognize, like, every other, you know, every other company like us must be having the same problem. Uh, so let's go down and work on that problem instead.

AI assessment note: “We launched the company in May of 2015.”

Q know, what prompts someone like you, I mean, everything we know publicly about a genre, I mean, when you left in 22, this thing is, it's doing great, you know, it's in the dev tech space, you know, it's, it's, it's grown like a weed. Wayne, why, why decide to sort of leave? Were you maybe just bored or something else? Can you share your story and transition to Tembo?

A Yeah, I mean, I tell a lot of people this, especially in the entrepreneurial ecosystem. I was, I was, I was offered a chance to sell my shares in secondary at our Series B, Series C rounds. You know, that was, this was the Zerp era, and I said yes. You know, I was basically, the way I saw it was, you know, I'm a founder, you know, sitting at a, at a poker table with a big giant pile of chips, and I have no, no money in my bank account, you know, and someone says, would you like to leave the table? It's like, Uh, yeah, let's, let's make this, um, equity real. Uh, so I had a rare chance of making it happen without, without the company being acquired. Um, you know, we just had a lot of interest in our, in our rounds and, you know, I'd been doing that, that company since 2015. So, you know, seven years in, um, I also had hired a CEO, you know, we hired a CTO, we hired a bunch of leadership. And so, you know, I just wasn't, I'm, I'm an early stage person. Um, I'm a from zero to blank, uh, person, you know, not, not what they're doing now. It's not my, not my talent.

AI assessment note: “I was offered a chance to sell my shares in secondary at our Series B”

Q Can you give us two or just so that the non-engineering audience can follow along here? Can you give us two or three of like the most popular one-liner prompts that folks might use? For example, take linear card and push it via GitHub or whatever. Can you give me three examples?

A I mean, imagine you have a corporate blog and, and you just shipped a new feature. You could say, you know, write me a blog post about this new feature. Look at this, uh, repo to understand what the feature is. So like a marketer could literally, uh, Tell it to inspect the, uh, the source code to find out all the things that just happened in the code base, and the AI is really, really good at that sort of thing. So, You know, it'll just generate the blog post, create a pull request in your GitHub, and, and, you know, you, you can, you can tweak it from there. You can actually just prompt it and say, hey, can you go a little more into this, and, and remove this piece, and, and so on. So, um, you know, it's very, like I said, it's very similar to what if you just had a software developer sitting next to you, and they will just do anything you want, whenever you want. Um, that's the feeling it is. So, so yeah, we use it to generate You know, if I found a misspelling on this webpage that you have pulled up right now, I could just prompt Tempo to say fixed, you know, linear is misspelled, and, and it'll just, like, find that in the website and fix it. Yeah.

AI assessment note: “write me a blog post about this new feature. Look at this, uh, repo”

Q Interesting. Okay, but what's your vision though? SMB, mid-market, enterprise, do you think your average customer's gonna be paying a thousand bucks a month going forward, a twenty million a year?

A No, we'll, we'll have million dollar a year customers. We have, um, Fortune 50 prospect right now that, you know, they have nothing like this. There's no product really like this in the enterprise space. Um, there are tools for engineers. That's what the current state, but there's no tools inside of, say, a Fortune 50 company for Anybody in the company to, to find any typo anywhere on any of their web stuff, and they could just like type, type a message to, to an agent, and it'll figure out the right repo, create a PR, get it to the right person to approve it. Yeah. Without, without having to know the ins and outs. So I think it's a huge, um, basically the more developers, the more people you have, the more powerful this tool gets. Mm-hmm.

AI assessment note: “No, we'll, we'll have million dollar a year customers. We have, Fortune 50 prospect”

Q Can you give us two or just so that the non-engineering audience can follow along here? Can you give us two or three of like the most popular one-liner prompts that folks might use? For example, take linear card and push it via GitHub or whatever. Can you give me three examples?

A I mean, imagine you have a corporate blog and, and you just shipped a new feature. You could say, you know, write me a blog post about this new feature. Look at this, uh, repo to understand what the feature is. So like a marketer could literally, uh, Tell it to inspect the, uh, the source code to find out all the things that just happened in the code base, and the AI is really, really good at that sort of thing. So, You know, it'll just generate the blog post, create a pull request in your GitHub, and, and, you know, you, you can, you can tweak it from there. You can actually just prompt it and say, hey, can you go a little more into this, and, and remove this piece, and, and so on. So, um, you know, it's very, like I said, it's very similar to what if you just had a software developer sitting next to you, and they will just do anything you want, whenever you want. Um, that's the feeling it is. So, so yeah, we use it to generate You know, if I found a misspelling on this webpage that you have pulled up right now, I could just prompt Tempo to say fixed, you know, linear is misspelled, and, and it'll just, like, find that in the website and fix it. Yeah.

AI assessment note: “write me a blog post about this new feature. Look at this, uh, repo”

Q Interesting. Okay, but what's your vision though? SMB, mid-market, enterprise, do you think your average customer's gonna be paying a thousand bucks a month going forward, a twenty million a year?

A No, we'll, we'll have million dollar a year customers. We have, um, Fortune 50 prospect right now that, you know, they have nothing like this. There's no product really like this in the enterprise space. Um, there are tools for engineers. That's what the current state, but there's no tools inside of, say, a Fortune 50 company for Anybody in the company to, to find any typo anywhere on any of their web stuff, and they could just like type, type a message to, to an agent, and it'll figure out the right repo, create a PR, get it to the right person to approve it. Yeah. Without, without having to know the ins and outs. So I think it's a huge, um, basically the more developers, the more people you have, the more powerful this tool gets. Mm-hmm.

AI assessment note: “No, we'll, we'll have million dollar a year customers.”

Q Yeah. I don't understand. So what did you pivot from? That was what you were. What are you now? How'd you pivot from that?

A So, so what we were was a company that wanted you to send your data to us, your clickstream data, you know, like user clicking on your website sort of data. So, um, that's what we needed in order for our product to make sense. And we were just doing like cohort, um, retention analysis and how, you know, showing how cohorts revenue grow and all, you know, all that cool stuff. Um, we could get customers to say, yes, we can get them to give us money and took us usually like four to six weeks of nagging to get the data. So, um, We basically decided, like, let's help these companies get data to services like the one we had, and that's a bigger, more global problem.

AI assessment note: “We basically decided, like, let's help these companies get data to services”

Q Very good. That was a mouthful. What a big problem. What's, what's astronomer do?

A Oh, so you didn't understand any of that, huh? Damn it. No, no, we're basically helping companies, uh, figure out how to use data to benefit their business. We work with later stage startups. We work with some big companies too, and really everyone's struggling. Um, and well, there's an opportunity. I'd say no one's really struggling, but everyone wants to figure out how to use data to make their business better. Uh, there are a lot of great products, you know, that have merged over the past few years, but, um, Well, I think like mixed panel, there's amplitude in the, in the, in the web analytics space. Google analytics is, is obviously pretty dominant. Um, but you know, with the rise of data science, um, a lot of companies are trying to get ahold of their raw data so that they can do more interesting things with it. And so that's really where we come into play to help them get the raw data from all these SAS silos or, you know, inside companies, oftentimes there's a lot of different databases where data resides and, uh, the data scientists need it all to be together to, to do their work.

AI assessment note: “help them get the raw data from all these SAS silos”

Q So is that the best analog here is the pricing page we're looking at right now? If I'm paying 200, I should assume kind of like that's 200 pull requests I'm getting.

A No, I would say it's more like, I would say it's more like, I like to say an average of five credits. It's on the high side, but sometimes you ask it to do something that might spend 20 credits. So we don't know, like, you know, these LLMs are sort of mysterious. Um, you ask it to do something hard, Uh, we use cloud code inside. Maybe you've heard of cloud code and, you know, cloud code will keep working at solving the problem until, until it's done. And so, uh, it's, to me, the more, the more money you're spending on Tembo, uh, the more value you're getting out of your engineering team because you're getting them work to approve, you know? So, um, yeah, we, a lot of companies are having this attitude with their engineers. They, uh, Winning means spending lots of credits on engineering, not less credits.

AI assessment note: “No, I would say it's more like, I like to say an average of five credits.”

Q So is that the best analog here is the pricing page we're looking at right now? If I'm paying 200, I should assume kind of like that's 200 pull requests I'm getting.

A No, I would say it's more like, I would say it's more like, I like to say an average of five credits. It's on the high side, but sometimes you ask it to do something that might spend 20 credits. So we don't know, like, you know, these LLMs are sort of mysterious. Um, you ask it to do something hard, Uh, we use cloud code inside. Maybe you've heard of cloud code and, you know, cloud code will keep working at solving the problem until, until it's done. And so, uh, it's, to me, the more, the more money you're spending on Tembo, uh, the more value you're getting out of your engineering team because you're getting them work to approve, you know? So, um, yeah, we, a lot of companies are having this attitude with their engineers. They, uh, Winning means spending lots of credits on engineering, not less credits.

AI assessment note: “No, I would say it's more like, I like to say an average of five credits.”

Q 30 have converted to paid, and together they've had a thousand merged PRs, which is your definition of that value loop. What, what, what usage metric do you want your users to say, where you go, man, they're going to be a long-term customer, I can upsell their whole company, We can get from one engineer to a thousand engineers. Is it one merged PR per, like, what is it?

A Oh, yeah, that's a good question. I don't, I don't have the answer to that yet, but I definitely think it's more than one. Um, you know, I think that, you know, if you did 10, for example, so we give away 50 credits, by the way, when you sign up. So that's the equivalent of about 10 PRs worth of, uh, credits. Um, so I think after about 10, you're gonna find a couple that you were like, oh, man, that was a game changer. Um, and, you know, we hope that Yeah, there's so many tools to experiment and play with right now, but, you know, we're trying to just build the most dead easy, simple user experience that people will come back to, embedding it in their own tools, so you can, from linear, you can assign work to Tembo, from Jira, you can assign work to Tembo, so yeah, we're, we're just trying to make it, uh, as just natural as, as, uh, asking a real developer to do, to do work for you.

AI assessment note: “I don't have the answer to that yet, but I definitely think it's more than one.”

Q So you're going to offset that, you think?

A Yeah, yeah. So at scale, we're going to be selling software. Um, uh, how do we, you know, again, um, I think we could price it based on value too. I think I can get a fortune 500 to pay me the equivalent of one developer if I can amplify a hundred developers, uh, you know, by 10%, if not more. So, so yeah, I think we can get value-based pricing on our product. And, um, now, you know, the other risky thing in this space is it's so easy to make software. It's so easy to knock off someone else's product. So, so I think like, yeah, so I think GTM is like the new, the new moat, you know, like if I had a squad of, you know, 200 sellers, uh, selling our product, I don't care how good your product is. It's hard to, hard to beat us, you know, if they're, you know, I think we're going to have a lot more, um, Um, you know, like to me, it's like the equivalent of like restaurants or car brands, you know, like they're all cars. They all look a little different. They're all pretty good.

AI assessment note: “Yeah, yeah. So at scale, we're going to be selling software.”

Q of 80, 90, a hundred million ARR, and you like are talking with such excitement about, man, I'm, I'm looking at my 30 users every day, people are getting addicted, I'm getting excited, this is like so different, this is like one 100th of what the largest enterprise customer an astronomer is probably paying. Um, what are you sort of obsessing over in these early days? This is good stuff.

A Yeah, well, I mean, zero, you know, to get to a hundred million in revenue, you have to pass through 10, you know, you have to pass through a hundred, you have to pass through a thousand. So, um, this is what early stage is all about, you know, building something that people want. You're doing it quickly, you know, in two months to get to build something that people will pay for is, is a hard thing. Uh, you know, we're in a very competitive space again. Uh, but, uh, yeah, I don't know. This is just, this is what, what is fun. If you, if you don't enjoy this stage, uh, This is the good old days. I always like to say, we knew it going in through an astronomer, like the good old days are now, um, everything's gets worse from here in terms of like, uh, at least for me, that's probably my own bias, but I, you know, I'm not a big, big company fan. So, uh, this is the, this is the best time I could have.

AI assessment note: “building something that people want. You're doing it quickly”

Q 30 have converted to paid, and together they've had a thousand merged PRs, which is your definition of that value loop. What, what, what usage metric do you want your users to say, where you go, man, they're going to be a long-term customer, I can upsell their whole company, We can get from one engineer to a thousand engineers. Is it one merged PR per, like, what is it?

A Oh, yeah, that's a good question. I don't, I don't have the answer to that yet, but I definitely think it's more than one. Um, you know, I think that, you know, if you did 10, for example, so we give away 50 credits, by the way, when you sign up. So that's the equivalent of about 10 PRs worth of, uh, credits. Um, so I think after about 10, you're gonna find a couple that you were like, oh, man, that was a game changer. Um, and, you know, we hope that Yeah, there's so many tools to experiment and play with right now, but, you know, we're trying to just build the most dead easy, simple user experience that people will come back to, embedding it in their own tools, so you can, from linear, you can assign work to Tembo, from Jira, you can assign work to Tembo, so yeah, we're, we're just trying to make it, uh, as just natural as, as, uh, asking a real developer to do, to do work for you.

AI assessment note: “I don't have the answer to that yet, but I definitely think it's more than one”

Q of 80, 90, a hundred million ARR, and you like are talking with such excitement about, man, I'm, I'm looking at my 30 users every day, people are getting addicted, I'm getting excited, this is like so different, this is like one 100th of what the largest enterprise customer an astronomer is probably paying. Um, what are you sort of obsessing over in these early days? This is good stuff.

A Yeah, well, I mean, zero, you know, to get to a hundred million in revenue, you have to pass through 10, you know, you have to pass through a hundred, you have to pass through a thousand. So, um, this is what early stage is all about, you know, building something that people want. You're doing it quickly, you know, in two months to get to build something that people will pay for is, is a hard thing. Uh, you know, we're in a very competitive space again. Uh, but, uh, yeah, I don't know. This is just, this is what, what is fun. If you, if you don't enjoy this stage, uh, This is the good old days. I always like to say, we knew it going in through an astronomer, like the good old days are now, um, everything's gets worse from here in terms of like, uh, at least for me, that's probably my own bias, but I, you know, I'm not a big, big company fan. So, uh, this is the, this is the best time I could have.

AI assessment note: “building something that people want. You're doing it quickly... build something that people will pay for”

Q Interesting. And what, give me a range. What kind of valuation are you trying to aim for?

A That's a great question. Whatever we can get it at, uh, to be honest, is it's, uh, you know, the, um, uh, great advice, you know, is, is think about your, your share price instead of your percent ownership. And, you know, if we can get this round done, our share price, our value of our equity rises and, and our, our, the odds of us living, uh, goes up. But, uh, at this point, like, you know, A's are tough, um, still. So we don't really want to bump our valuation up too much. So it's going to, we're, we're going to just Listen to what the offers are and, and not be too greedy on that. Uh, so we have some, some, some head, some headroom.

AI assessment note: “Whatever we can get it at, uh, to be honest”