Key Takeaways

  • Public private equity firms are holding intake classes flat or cutting junior analyst and associate seats as software takes over core analytical tasks.
  • Intapp President of Industries Ben Harrison reports that financial models and pitch decks that once took days or weeks can now be completed in hours or minutes.
  • Private equity managers will capture direct margin expansion by reducing the junior labor traditionally required to process transaction data.
  • Multiple Intapp clients are running AI agents inside investment committees to test recommendations and evaluate deal assumptions.
  • Human review remains necessary because automated outputs require iteration and judgment before investment partners commit capital.

The Analyst Pyramid Is Compressing

When Ben Harrison worked as a junior investment analyst, the job required long nights assembling financial models and pitch decks from scratch. “Knowing that it took me days and weeks to build models and decks, being able to do that now in hours or minutes, you still gotta digest it and iterate it, it's not perfect,” Harrison told Ted Seides. The mechanical work that once justified hiring large cohorts of twenty-two-year-olds from investment banking programs has compressed into software workflows.

That compression is now altering hiring plans across large alternative asset managers. “You're starting to watch the big public firms at least hold constant new analyst and associate classes, maybe trim them some,” Harrison noted. For decades, private equity relied on an intake pyramid where dozens of junior staff processed raw data so partners could execute two or three buyouts a year. As frontier models automate those workflows, sponsors do not need the same volume of junior headcount. The immediate financial result is straightforward: “There's gonna be some major margin pick up in the junior ranks that we've been using in these industries to process data and information.”

AI Agents Move Into the Investment Committee

Automation is also reaching the partner level. Harrison pointed out that sophisticated private equity sponsors are already testing software directly inside their primary decision meetings. “We've got a lot of clients who have AI agents sitting in investment committee to see what it can do and to see what recommendations it makes,” Harrison said.

These systems do not have voting rights. Harrison made that distinction clear: “Is the investment committee gonna be an AI agent? I don't think so, but are they gonna use it to help? For sure.” Instead of replacing senior partners, these agents act as an automated check against deal teams. They review underwriting assumptions, compare metrics against historical firm data, and point out risks that deal champions might skip over when trying to push a transaction forward.

The strategic tension sits in talent development. If associates no longer construct financial models line by line, they miss the repetitive analytical work that builds investment judgment. Harrison observed that automated models are "not perfect" and require human review. As software handles the baseline data assembly, private equity firms face a new question: how will the next generation of investment partners learn to evaluate deals if the junior analyst desk no longer exists?

Why It Matters

Management company economics in private equity will decouple from headcount growth, expanding operating margins for firms that automate data processing. Because modeling speed is now table stakes, competitive advantage will shift entirely to proprietary deal access, relationship networks, and executive judgment.