Our CEO Alex Robinson recently joined Chief Relationship Officer Brandon Sedloff on The Distribution by Juniper Square for a robust discussion on how fast AI is moving through private markets.
Here's a recap of their conversation:
AI crossed a real threshold last December
We crossed a threshold around December 2025, when the models became capable enough to automate more than 90% of the knowledge work in the economy, at Juniper Square and at the GPs we serve. That shifts the hard work from the intelligence layer, which the foundation labs control, to the applied AI layer: getting the technology to work for the particulars of your business. It's what I've spent the last eight months on.
We built an operating system for AI legibility
Making that shift starts with making your company legible to AI. Juniper Square has operated as an on-the-record company for years, well before the AI moment: every meeting, candidate interview, Slack message, and email is recorded. This spring, I layered a new operating system on top of that called Orca, short for Orchestration, Review, and Clearinghouse for AI. I now hold a daily global meeting with the top 50 to 100 people across the company, every workstream has a named owner and a connected repository, and if you miss a meeting, you go ask your AI what happened. It's led to a real speed-up in how we run the company.
Trust is still the first ingredient
None of that changes what I think is the most important ingredient in a GP relationship: trust. We work with more than 2,500 GPs worldwide, built spreadsheet by spreadsheet and fund by fund over twelve or thirteen years. That's the thing that really matters.
Three ways we're putting AI to work for GPs
We think about our applied AI work in three areas: the software we build, the fund administration work we deliver, and the technology platform GPs use to run their own businesses. All three already have products behind them. We shipped Headless GPX in June, so an agent, either one we build or one a customer builds, can connect to Juniper Square and get answers or do the work directly. We shipped Faye in July, which gives CFOs and fund controllers an oversight agent for their third-party fund administrator deliverables. And we're building a third effort called Agency, an agent development shop we'll launch publicly later this year.
We're a tech company that became an administrator
That range of products is also what separates us from a traditional fund administrator. We became an administrator six or seven years ago and now operate at real global scale, but we started as, and remain, a technology company. Most administrators are primarily human capital services businesses. The future calls for having AI do the work, with humans steering the outcomes and taking ultimate accountability. That distinction, a technology company versus a human capital shop, is what drives everything else for us.
The market is splitting into sharpshooters and scalers
Looking further out, I see the GP landscape splitting into two camps. One is what I call sharpshooters: managers with a distinctive, non-repeatable strategy that's inherently limited in scale. Think of a fund that only invests in whiskey barrel futures, but can still produce 50% IRRs and 3x MOICs within that narrow lane. The other camp is going all in on scale, the way Blackstone, Apollo, KKR, Brookfield, and Carlyle already have. As AI lowers the cost of having a good idea, I expect more managers to become sharpshooters, with platforms like Juniper Square handling everything else. You're going to have to pick a lane. It's going to be really hard to sit in the middle.
For the full conversation, including Alex's advice to GPs preparing their own firms for AI and where he sees private markets heading, watch or listen to the full episode of The Distribution.