Juniper Square raises $130M Series D and launches Juni AI to bring deterministic AI to private markets fund administration

were 5 minutes over. We are 11 minutes. We have a We have a light system in the studio, but uh the golden retriever does not concern himself with the lion does not concern himself with who's in the waiting room. Uh we have our next guest from Juniper Square. Thank you for waiting. Great to talk to you.

And we got some news. We got some news. We got some news. What do we got to be here, guys? What do we got? Uh, we got two things. We got a $130 million series D. Congratulations. I've been waiting to do that all day. Glad we got it out. Yeah. And then we launched our AI product that we call Juny AI. Congratulations.

That's our second uh bit of news. Now that we got the important stuff out of the way, who are you? What do you do? Yeah. Yeah. Well, I'm I've become a fan of what you guys are doing. Uh, so I'm glad to be on the show. It's really interesting and unique. So, keep it up. Um, but yeah. Yeah. So, my name's Alex.

I'm the co-founder CEO of Juniper Square. Nice. And, uh, Juniper Square is an 11-year-old technology and services company. All of our customers are uh private markets GPS. Sure. Uh so venture capital firms, private equity firms, real estate investors, crypto, private credit, you name it.

If they raise an outside fund, Yep. Then we help them with fund administration, we help them with technology to manage that fund, reporting, basically everything is not the investing. And we've got close to 200 customers now, trillion dollars that we're responsible for stewarding and about 750 employees. 2500 customers.

Is that is that not all of them? How many investors are there? I feel like I saw that number on your website. You need to update it, by the way. It says 2,000 on your website here. 2500. Yeah, we got to get that get those numbers out.

But I mean I I feel like there's I mean we try and talk to I thought we talked to all the VCs. We've had we've done hundreds of interviews. I thought we talked to them all. Apparently there are more. There's some left. How big is this market?

It it feels like like you know you talk about uh uh private markets investing that feels like a small market yet series D is at a billion.

1 like you're building a big business how big is this market right so the you have to kind of take it by asset class right because VC is relatively small sure as an asset class both in terms of the capital that it manages and the number of uh GPS or or investment firms that are out there private equity and real estate are respectively probably five times as big as venture.

So there's thousands and thousands of, you know, just think of all the middle market private equity kind of LBO growth focused uh firms out there or think of all the different real estate investors focused on, you know, apartment buildings or retail or whatever. U there's probably about 20,000 GPS globally. Mhm.

Uh that matter. Was Zer crash hard for you? Was that a slower growth period for your business? Because it seemed like it was really rough for all the VCs and a lot of investors pulled back, but it's not as like they really shut down their funds. Yeah, it's it's more like net new fund creation. So, yeah.

Was that a real driver or are you just going after the existing players and so the new ones don't necessarily matter as much? No, we definitely felt it and you're right that we're what we're sensitized to is in a given quarter, a given year, how many new funds were created. Mhm.

Our market share is so low that we still can sell to existing funds and people switch their providers all the time, but it's always easiest to get the new fund that's raised. So, we had the double whammy of uh you know, big focus for us is commercial real estate.

So, when COVID hit in spring of 2020, you know, it just took out retail, it took out hospitality and hotel and commercial real estate was it was really tough sledding from like 2020, you know, really until 2022. And then the rate hike started and commercial real estate's, you know, heavily capex intensive.

It's super rate sensitive. Um, and then we had the big bust in, you know, public markets correction. We had the the big overhang of everyone investing at, you know, 100x revenue multiples in 2021, 2022. We're going to bring those back. Yeah, we're working we're working on it here.

The sober the sober days of uh 100x revenue multiple, which we're kind of back to now areas. Uh how did you get through all those times? It's recovering. It's coming back. Uh h how did you get through the tough times? Did you have to do layoffs or change the direction of the business? Were there any pivots in the story?

Kind of specific meditation tracking meditation actually. Okay, cool. Yeah, bring it down. Yeah, all of the above you guys. We we had to do some layoffs for sure like everybody. I mean, we were going into, you know, the Fed started hiking rates in like late 21. But by spring of 22 is when we had that huge correction.

Yeah. In the public markets, and by that summer, it was clear that companies like us and pretty much every high growth late stage tech company was spending too much, right? We were uh all investing for revenue growth and we all had to pivot to invest for growth efficiency.

Um, so we did a round of layoffs then like a lot of people did. It was the first time I ever to do that as an entrepreneur. It's not fun by any means, but it was necessary.

Uh, but then the big thing for us is, you know, we started out as a technology company, but then we added uh in 2019, we moved into the world of fund administration and we became a full-ervice fund administrator. And fund administration is like a really critical market for private capital.

There's probably 3040 trillion dollars of private capital out there in the world. and all almost all of it needs a third party administer administrator.

So we we kind of moved into this market that was like 50 times the TAM of our previous market concurrent with this big correction happening and so those forces ended up working out that like the company did really well and we continued you know we grow grown really quickly uh through a really challenging market period.

Yeah. Um so it wasn't without its bumps and bruises but it's overall it's been really well. Yeah. What? Talk about uh the new AI product. Uh I imagine it's a difficult one to build.

You can't really like LPs and investors can't really tolerate hallucinations, you know, and uh I can imagine it it you know, a product that works well. I I don't mind if you leak my fund returns to my direct competitor. No problem. Well, well, no less less of that.

But if you're trying to understand fund performance and things like that or reporting, you can't like, oh, I'm sorry. I misreported TVPI. Yeah. Yeah. Yeah.

So that is the crux of the challenge is um I mean as you guys know you talk about it a lot on the show you know generative AI is probabilistic right so next token prediction is is probabilistic in its nature and the kind of work that we do of reporting on financial assets is deterministic right like 1 plus one always has to equal two.

It can never equal you know one or three or anything like that. Um, and so the real trick is, uh, and as if you guys ever know, if you've ever like taken a pile of invoices or something and uploaded them to chat GPT or any LLM and been like, "Hey, tell me the, you know, invoice amount for invoice number eight.

" It's it's terrible at financial data extraction.

Um, and so the big hole that we have in the market that we're filling for our customers is if you're an asset manager, if you're a financial services firm like a venture capitalist or real estate investor or whatever, one of the key things you have to do is you have to report to your regulators, you have to report to your LPs, that reporting is like, you know, you're putting your brand and reputation and your ability to raise more money in the future and everything behind it being accurate and correct.

Um, and so the trick is how do you leverage all of the power of generative AI? This like all these summation use cases, document extraction, the chat interface that we've all become so accustomed to.

How do you fuse that with a set of deterministic tools that do all the precision stuff, the precision math, so that like you're calling functions in Python to run some code to, you know, answer a question. You're not trying to have a next token prediction be the answer of that.

And that's basically what this platform is is um it's multimodel. We have a bunch of different models running inside of it. Uh uh and the idea is think of it as like a model orchestration layer where if you're a private markets GP, there's a lot of work you got to get done. You want agents for different things.

Different models fit different use cases. Um and you need a way of orchestrating all of that work, ensuring it's compliant, ensuring it's accurate. Uh and that's what this layer does. Mhm. Anything else? This was great. Yeah, we Yeah, we got to have you back on.

We're running behind today as you saw, but we really appreciate you stopping by. We're congratulations on on making it through the fire and uh come back on when you have more news. Yeah, we'll talk soon. Happy to be here, guys. You and the team. Cheers, Alex. Bye. Um, really quickly, how'd you sleep last night?

I'm hopeful. I'm on a run. 94. Let's hear if you beat me last night. I'm going to be streaming 86. Oh, let's go for John. It's crazy. Ever ever ever