The finding
KPMG describes private equity firms moving from testing AI to operational deployment, with agentic workflows tied to value creation plans. The EBITDA case rests on reworking high-frequency, repeatable processes in finance, HR, IT, sales and customer support.
The piece names clean, governed data as a prerequisite, and it says people and change management matter as much as the technology. Sharing playbooks across a portfolio speeds adoption.
Source: From AI pilots to private equity portfolio performance, KPMG LLP, from an interview in Private Equity International, 2026.
What it means for you
For an operating partner, this moves the question from which tool to try to which process to change. It also sets a higher bar: the result has to show up against a baseline.
It favors a repeatable approach. A process fixed at one portfolio company can become the playbook for the next.
Our take
We agree with the order of work: process first, then data, then the model. Our method sorts each step into delete, code, agent or human review, because many steps do not need a model at all.
The piece is qualitative and gives no figures, so it will not size your opportunity. Pick one high-frequency process at one company, and agree on the baseline before anyone builds.