We’ve had 47 calls with private equity teams since May. Operating partners, value-creation leads, portco CEOs and CTOs. Here’s what they’re saying about AI and how responsibilities are changing inside a portco: ONE - People are the bottleneck, and they bring it up first. Adoption and change management came up as the bottleneck in 16 of the 47 calls. One fund’s AI lead put it this way: “Change management is the name of the game a lot of times with these things... That’s a big reason why our team has grown.” They grew the team to put people on site at portcos and help staff change how they work. TWO - Teams are too big, but nobody wants to cut without proof. A value-creation lead looking at a portco with about 500 engineers told us: “the teams are too big and I’m advocating for smaller teams... I can’t do that without showing that there is improvement in productivity first.” So the first project is to measure what changes on one team. The reorg comes later, if the numbers support it. THREE - What happens to the freed-up capacity is still an open question. On the same call: “Whether that’s headcount reduction or that’s improved throughput and therefore better products. We can discuss where that productivity goes.” A CTO on another call had already made his choice: “I would probably rather triple our output than scale down.” Before a portco starts with AI, someone needs to answer: → Fewer people, or more output from the same people? → Which team goes first? → What number proves it worked? → Who owns the AI after the pilot? → What does it cost to run at full usage? FOUR - Buy-in decides who gets help first. “I am less keen to engage any portfolio company where the leadership isn’t as bought in.” The CEO of one portco in that fund has drafted a note to staff but hasn’t sent it yet. The message is roughly: “either you’re adopting and using AI and going faster or you’re not going to be part of our growth.” FIVE - My biggest takeaway: everyone’s role is changing at the same time. The operating partner has to prove the productivity gains before anyone starts restructuring. The CEO has to make it clear that using AI is part of the job. The CTO has to work out how to run smaller teams with less ceremony. “At my last company we let go of a lot of the scrum structures and it was fine,” one SVP of Engineering told us. A PE-owned distributor now has a “digital AI asset manager” doing the day-to-day work of a product owner. Everyone has work to do. If you’re an operating partner, pick one team in one portco. Before changing anything, measure rework, cost per finished task, and time from ticket to production. Answer the five questions above. Then run a small AI-assisted team alongside it for a month and compare the results. That’s what we do inside portcos. We start with one workflow and measure the results against a number the business already tracks. We work month to month, and if we don’t deliver, you don’t pay.