Versori Brings Agentic Integration to Avalara’s Ecosystem

Hey everyone. We're wrapping up our Avalara Crush North America coverage. My next guest is Sean Brown.

Yep. And Sean joins us from the UK, though he's here obviously with me now. Well, first of all, let's welcome him.

Sean, welcome to Techstrong TV. It's great to have you on here, man. Thank you very much for having me.

My pleasure. It's an absolute pleasure. So Sean, you know people always want to know who they're watching here.

Tell them a little bit about you. Well, compared to some of the guests that you've had on, it's nothing to note, but I grew up in a small town called Coventry. Sure.

Single mom. Didn't really have much of a success in education, et cetera. So when I was around 18, I went out to Australia.

I was about to play soccer at a good level. Okay. And tried to make it as a professional in Australia.

Failed massively. And then, ended up training to become a diving instructor, a scuba diving instructor. That's quite cool.

Yeah, you didn't think you'd go this way to do this, but- In Australia, too. No, I didn't ... so then I ended up working my way up onto big superyachts.

Uh-huh. And I taught Leonardo DiCaprio to scuba dive. Really?

Kate Hudson. Oh my God. That's great.

Loads of celebrities. Yeah. And then, so while we're at sea, and I was working for these huge billionaires that have great businesses, and because I teach them to dive, you get to know them a bit better than just general crew.

Sure. And we had terrible internet, and I wanted to build something, so I started trying to build my own website, and it was a social media platform for scuba divers to share- Really cool ... their dive experiences and build a bit of a community.

But then Facebook Pages came out and kind of destroyed that as a- Yeah. They- ... very niche market.

Yeah. It was. Not a billion-dollar market cap there.

You learn a lesson, though. But yeah, I taught myself to code. " So, developed an influencer marketing platform, and then we sold that to a company up in Manchester.

Mm. And actually, the guy who was managing director of that is one of the biggest podcasters in the world now, Steven Bartlett. Really?

That's cool. He recently bought Visori as CEO. So I worked with his site, close team.

We grew that business once I've sold into them. And we're a social media company that had lots of different channels and influence, so I started building the commerce element to it. And that led me to see that there was a big issue with drop shipping and marketplace selling.

So I founded a company called Mercato, and that really shot out. We didn't do any fundraising or things, but we scaled to $10 million ARR in 18 months- Wow ... completely bootstrapped.

And we're running the marketplace solution for some of the largest retailers in the world. And then I started finding an issue. All of our engineers were working on integrations nonstop.

Yeah. We looked at the market for integration platforms like the Boomi, the MuleSofts, et cetera, and nothing fit the bill because we wanted to do something slightly custom outside of the out-of-the-box connectors. So we ended up building them ourselves.

But it inspired me to think that I didn't really want to build a professional services organization, is what that company was turning into. So when I exited that, I decided to found Visori, and this is where we are today with Visori. We went through Y Combinator for Visori.

Oh, okay. Out in San Francisco? In San Francisco, yeah.

Okay. We moved to San Francisco just as my firstborn was born. Wife loved it.

Why not? So had to move out to San Francisco. We went through Y Combinator, and then, yeah, we decided to start building for the future of integrations.

Very whistle-stop tour around that. That is. But what a charm story, man.

What a charm story. Good for you. So I get now why you did Visori.

When did AI kind of rear its head in there? Yeah, so from the very beginning, actually. So when trying to figure out how to do integrations, GPT-2 was out and not very well-known.

5 that really got going. Right. That was the one that broke through.

But I was playing with GPT-2, and then I was building these kind of crutch system prompts that would help it write code. And I saw that AI was going to write code pretty soon. Mm-hmm.

And they're only going to improve these models. So when I stood out of what the legacy IPAS integration platforms had done in the market, I was like, "Look, I want to be naive on purpose to go and build this. " Mm-hmm.

" So we started building the infrastructure side of integration, so all the things like authentication, credentials handling, key value store, request response, back off, retries, assertions, all the really technical infrastructure layer of integration we started building for. But then also these engines that in the future-- They weren't called agents back then, but we were kind of building these sort of a planning agent, a connecting agent, and then a building agent to write the code. So we built this product and then the platform for the infrastructure in preparation, so this is like four and a half years ago, in preparation for where AI will be to today.

So we had this really audacious goal to go after a massive gamble, massive risk. We had some fantastic VCs that backed us in the early days after Y Combinator that believed in our vision. And now, over the last year, the models have become good enough for that vision to really become a reality.

So because we were already there, we've not had to be building to catch up. Right. And we're native in AI, so we are really the market leader.

You believed where it was going before it was there. I guessed, mate. I didn't believe.

I just guessed. It was a full guess. Well, because an old saying, so you played football, soccer, as we call it here.

Yeah. But in American football, they try to teach quarterbacks, "Don't throw the ball where the receiver is. " Exactly.

Love that. And it takes ages. And that's what you did here with the AI, right?

You didn't build it based upon two-point chat GPT-2. Exactly. I built it on- I wasn't going to do it ...

Claude Opus or Claude Mabel- Yeah ... and where we'll be able to get to now. So, yeah, super enjoyable.

And then, we built the platform for around about two, three years, and then just last year, we started opening the doors to customers, and we won huge opportunities with the likes of Walmart, Uber were using us, and lots of large technology platforms were using us as well, including Avalara. Ah. And that was our introduction to Avalara.

So they were a customer? They were a customer of Visori. So, okay, let's fast-forward.

How did you become part of Avalara, then? So funnily enough, I genuinely really enjoyed working with the team on it. And so sometimes you work with technology teams that say, "Right, we need to connect data B," and then they say, like, "You do it.

Put a prompt in," and then all the bats. Avalara's team always came with really good requirements and documentation, and we just thought, "Oh, this is an enterprise that has their- Their act together ... " I was going to say a different word there, but- No, I believed you.

Good. And then, so really enjoyed working with the team. We started delivering integration super quick that would take months and months, we're doing it in days.

Mm-hmm. " And that's really interesting because our saying is we want to be part of every transaction in the world. Transaction.

" And then we did this playful exercise of analysis of our target market and their customers. And every single customer in Avalara- Overlap ... overlap where they need integrations, and either into Avalara or beyond.

And then I started saying, "Huh, you can do a lot of introductions for us," et cetera. And I'm a builder. I love building technology.

I'm not a go-to-market guy. Mm-hmm. I'm not a good salesperson.

I can sell something I'm passionate about because I want to talk about it. Sure. That's what makes a founder.

I think so. And I'm not good at the go-to-market, the strategic side, and the execution. I would love to be, and hopefully I'm learning that while I'm at Avalara.

So when I saw this huge challenge, we're about to go through our next fundraising. I saw this huge challenge ahead of us is that how do we actually get to the next level of scale? And I did some maths and I thought, well, working with Avalara's go-to-market team, being even by osmosis, part of some of their deals, we could expedite our go to market by five years.

Even more. There's a few other acquirers, those other IPASSes that were trying to acquire us at the same time. Mm-hmm.

But funnily enough, they wanted to acquire us to kill our technology so they didn't cannibalize their own. Their own. Wanted us in the AI teams.

And I've done a lot of M&A over the years. During the dot coms, I was part of a company, we did 30 acquisitions in 36 months. Hmm.

It's almost one a month we were buying companies. Wow. What you're describing, right, so a lot of times companies make a build versus buy decision.

Yeah. Do I build it myself? Do I buy it- Yeah ...

instead? You did a reverse build versus buy. Do I sell it to someone who helps me get there, or do I take the long road hoping the market will be there when I finally arrive?

Exactly. Yeah. And, generally those are good...

That long road can be a very long road. It could be a road to nowhere. Yes, and that's the risk of being a founder, right?

Yeah. Fortunately, founders that like the zero to one, they feel safer in the risk- Yeah ... than they do in stable employment.

There are founders like that, right? Because they're just going to bet on themselves. Yeah, exactly.

But, I've sold my share of businesses, too, and I learned this from the first person, my first company that I sold, it was to a man who had done a lot of roll-ups. He did very well out of his life. Right.

Yeah. " Yeah. " And by the way, that's stuck with me.

Since I built that social media platform for scuba diving, we actually did sell it, but to a professional association of diving instructors, PADI. Have you ever heard of them? Hmm.

I'm PADI certified. Yeah, me too. That's awesome.

Oh, okay. Well, you're an instructor. I'm just- So when PADI came along, we wanted to build Scuba Earth.

Mm-hmm. They just bought the code, and I was like, from something zero, and I learned to code on a tiny web book. Uh-huh.

Really? It wasn't even a laptop. It was a little web book.

Okay. And that's all I could afford. It was something like £150 or something like that.

Uh-huh. And then when they said, "I want to pay for this thing," it was like 20 grand or something. " I was like, out of thin air, I've developed value that someone else wants.

Good, and someone's going to give you a nice check. And that's been my drug in life, is like creating something from zero- There is. It is ...

and giving it to other people. So- Once you... So I, look, I did about 20, 25 years in startups.

Yeah. Okay. Buying, selling, building.

It is a drug. Yeah. Right?

Yeah. " And that's actually how this whole thing started, from a blog. Yeah.

True. But you still started another startup. Yeah.

Well, I did, but it wasn't on purpose. Yeah, got it. The other ones were on purpose.

But anyway... But now here, let's fast-forward because we're going to run out of time. So now you become part of Avalara.

Yeah, so, we were just so aligned on so many things. Mm-hmm. The acquisition happened around about end of February this year.

And then it's funny, Hugo, the CEO of Avalara, I was like- Yeah ... find it tough with being at the big table with the big kids, and I just didn't quite get corporate energy, even the people who don't give you corporate energy, it's still processes and systems- It's a big company ... and HR.

Yeah. I hear you. I go through that right now, but go ahead.

So since Hugo's come in, he's been super incredible with aligning me and focusing me onto one thing that really opens up value. And, the initial intentions were, Versori was just going to be completely standalone. But then we started seeing different people from Avalara like, "Hey, we need you for this integration," and we started get more and more and more, and it started rolling.

So we've got this really clear strategy now that partners of Avalara can go through the Versori partner program, where they do a little bit of training, and then they can use the product themselves to build their integrations and partners. Their own integrations. And then we've got the go-to-market team, so the account manager teams, and then the net new business.

When they sign a customer, whether it be on an API deal or they've just got a system that doesn't have an out-of-the-box connector or doesn't quite fit with what's already there, they then use Versori to unlock that connectivity. And then we still have Versori as standalone. When these customers of Avalara use Versori for the Avalara to ERP build, for example- That's it ...

they go, "Oh, that was cool. " Right. "Could we build this process automation?

Can we use it for other-" So now they're using Versori outside of the Avalara box, even. Exactly. So now they're using Versori in a much extended fashion as well.

So it's naturally executing itself and becoming this really perpetually growing business. And growing out into a new... Yeah.

Yeah, so- What a great story, man ... really excited about it. I'm locked in, I'm here for the long run, and I'm just loving it as well.

I'm quite lucky. I love to hear it. People want to get more information about Versori, what's their best place?

com. Spell it for me. V-E-R-S-O-R-I.

And a fun fact of where Versori comes from is a unit vector, which is, think of a sphere and you take three points on the edge of the sphere, and the angle in the middle is a unit vector. And that's transformative. We have data transformation.

And it's also called a versal. Okay. I spoke an I on the end of it to make it sound like a startup.

Hey, man. You're killing it. Pleasure.

Thank you very much. Excellent. Versori.

You'll see it in the bottom third of your screens, but go check it out. Hey, we're going to wrap up here at Avalara Crush. It's been a great two days.

tv, Techstrong TV YouTube, or the OTT channel on I... I shouldn't say iPhone. On iOS, Android, Roku, Amazon, Apple, however you watch your videos.

But, in the meantime, you know what? You've seen what Avalara's doing. Follow them online.

Check it out. We'll be continuing our coverage of Avalara. But for now, this is Alan Shimel on Techstrong TV.

We're out.