Insights
DefinitionNovember 19, 20266 min read

What does an AI operating partner actually do?

The person a fund holds accountable for turning AI into measured value across its portfolio: thesis, proof, shipped software and a number the IC trusts.

An AI operating partner is the person a private equity fund holds accountable for turning AI into measured value across its portfolio. The role sits beside the deal team and the other operating partners. It decides where AI should be used, proves the case in one company, makes sure the work actually ships, and reports the result in numbers the fund already uses: EBITDA, margin, capacity and payback.

The short version: a CTO looks after one company’s technology. An AI operating partner looks after one fund’s AI return.

The five jobs of the role.

  • Set the thesis. Decide what AI means for this fund’s companies, which is usually a small number of repeatable plays, not a list of tools. Most mid-market portfolios have a handful of workflows, such as intake, quoting, scheduling, document handling and reporting, that show up again and again.
  • Assess the portfolio. Score each company on readiness and opportunity, then group companies by the pattern they share so one design can be built once and rolled out many times.
  • Prove the case. Pick one company and one workflow, take a baseline, write a business case that management and the deal partner both sign, and get it live.
  • Make sure it ships. Hold the build to a date and a metric. This is the part most programmes skip, and it is the part that decides whether the fund ever sees a return.
  • Measure and report. Roll every live use case into one page for the fund: adoption, hours saved, net benefit by type and payback, kept in separate columns all the way to the total.

How it differs from the roles around it.

  • Portfolio company CTO. Owns one company’s systems and engineering team. An AI operating partner works across companies and is measured on fund-level value, not uptime or delivery.
  • Digital or value creation operating partner. Often covers pricing, procurement, sales effectiveness and technology broadly. The AI role is narrower and deeper: it has to understand models, data and evals well enough to tell a real use case from a demo.
  • Strategy consultant. Delivers a diagnosis and a deck. An AI operating partner stays through the build and is judged on what goes live.
  • AI vendor or agency. Sells a product or build hours. An operating partner is on the fund’s side of the table and should be able to say no to a tool.

What good looks like after a year.

  • A written AI thesis the investment committee has seen and uses in new deal memos.
  • Every company scored, with a first wave chosen and a reason for every company left out.
  • Two or three use cases live with real volume, each with a measured baseline and payback.
  • At least one design rolled from the first company into a second, faster and cheaper than the first.
  • An AI section in the exit story that a buyer’s diligence team can check.

Signs a fund needs one.

  • Every portfolio company is running AI pilots, and none can show a number.
  • Limited partners or the investment committee keep asking for the fund’s AI story.
  • Deal teams want AI diligence on targets and have nobody to ask.
  • Two companies are paying separately to build the same thing.
  • A company is approaching exit and its AI claims would not survive a buyer’s questions.

Full-time, fractional, or project-based.

Large funds with many portfolio companies increasingly hire the role full-time. For lower and middle-market funds, a fractional AI operating partner is usually the better fit: a senior operator for a set number of days a month, starting with a portfolio assessment and one proven use case, then scaling the time with the number of companies in flight. A project-based start, such as a portfolio assessment or a single value sprint, is a sensible way to test the fit before committing to a longer arrangement.

What to look for when hiring.

  • Operator, not observer. Someone who has run product or technology inside PE or VC-backed companies and has shipped software, not only advised on it.
  • Fluent in the fund’s numbers. Can turn hours saved into EBITDA the way the CFO would, and knows the difference between capacity, cost removed and headcount.
  • Hands-on with AI. Has built and run AI systems in production, including evals, human approval steps and the cost of running them.
  • Willing to stop things. A ranked list of AI plays is only useful if it also names the ones to stop.

Common questions.

Is an AI operating partner the same as a Chief AI Officer?
No. A Chief AI Officer usually sits inside one company. An AI operating partner works for the fund across many companies and is measured on fund-level value.
Does the AI operating partner build the software?
They should be able to. At minimum they write the business case and hold the build to a date and a metric. In smaller portfolios the same person often leads the first build with the company’s own team.
Where should a new AI operating partner start?
With a portfolio AI assessment: score every company, group them by shared patterns, and pick one company and one workflow to prove first.

This is the role I play for funds, usually starting with a Portfolio AI Assessment and one proven use case. The Approach page sets out the full sequence.