It's Already in Your Portfolio
AI is running inside your portfolio companies right now, whether anyone signed off on it or not. In one healthcare business we worked with this quarter, roughly 95% of staff were already using personal AI accounts while the firm had no policy at all. That is closer to the norm than the exception. The decision is no longer whether to adopt AI. It is whether to get in front of what your teams are already doing.
Demand for AI advisory rose 193% year over year (YoY) in Q2 2026, and nearly every question we now field is about how to execute, not whether to start. The constraint is operational, not aspirational, and that plays to what your team already does best.
Sequence the Backlog
“What runs first against the value creation plan and the hold clock, and what do we kill?”
This was the single most common opening question. Firms already have the list; the ask is how to prioritize and cut, because you cannot chase everything at once. The work they want is a read tied to ROI and exit timing, not another inventory of what is possible.
What we’re hearingMost PE firms run parallel pilots and hope one sticks.
What the movers doThey sequence by ROI against the exit clock, run one or two, and cut the rest fast.
Grade What You Already Own
“Across the portfolio, where is AI a real threat to the thesis, where is it upside, and which companies are really moving versus telling me they're moving?”
The sharpest deal-side signal this quarter was a gap. Firms told us technical diligence is well covered; the strategic read is not. One firm underwriting a roughly $100M telehealth provider wanted a single clear answer: could emerging AI commoditize the target's edge, and where could it compound margin after close? Generalist diligence doesn't cut it. The same lens turns inward: build a standing AI impact scorecard across the book you already hold.
What we’re hearingMost PE firms grade portfolio AI exposure ad hoc, deal by deal.
What the movers doThey keep a standing read on AI exposure and revisit it every board cycle, like any other value driver.
Ship It. And Keep It Alive.
“How do we get from assessment to production, with a partner who builds rather than leaves a deck, and who owns it once they're gone?”
Getting to a working build is the easy part now. Sustaining it is where value is won or lost. The more common failure is quieter but deadlier: the thing gets built, and no one uses it. One building-products distributor had already invested in dashboards and reporting and still could not turn "here is a chart" into "here is what to do this week." Layering AI on top changes nothing until someone owns the last mile, turning output into decisions people act on.
What we’re hearingMost PE firms buy a pilot and find out later that nobody owns it.
What the movers doThey settle ownership before they build, and treat "who runs this in six months" as part of the scope.
Govern It So You Can Defend It
“What's the guardrail my portcos can operate inside, that I can stand behind with my LPs, without smothering adoption?”
The trigger is rarely a breach. It is a portfolio company already using AI in ways no one at the fund approved, and a partner asking what the firm's position is. Most firms don't have one yet. The firms getting this right set one standard the whole portfolio operates inside: what data can go where, which tools are sanctioned, what has to stay in-house. It travels across companies instead of being re-litigated at each one, and it is written so a partner can defend it to an LP without flinching. That is a fiduciary posture, not IT hygiene.
What we’re hearingMost PE firms react once a portfolio company is already exposed.
What the movers doThey set the expectation and make each portco stand up its own policy against it, so it holds across the portfolio and in front of an LP.
The Frontier, Honestly
“Past the demo, what's genuinely production-grade in a lower-middle-market company today?”
The bar is moving fast, and the gap between frontier and floor is widening. The most advanced buyers already treat last year's tools as dated. One built an agentic stack to run across sourcing and diligence. But the same buyer was blunt: none of it works until the data underneath is in order, and most companies aren't there yet. That is the honest answer on what's production-grade. The ceiling is high, the floor is data. Agentic and governance are now the same conversation: the further out you push, the more governance lets you scale safely. One caution worth keeping: AI is a thinking partner, not a thinking replacement.
What we’re hearingMost PE firms are piloting agents.
What the movers doThey fix data and governance first, because that is what makes agents safe to scale instead of a liability to unwind.
The Constraint Was Never the Technology
Across every stage of this work in Q2, the pattern held: the thing in short supply was not vision, it was capacity to execute and sustain. The slow deal market handed you a window most PE firms will waste. The ones converting it into operating leverage now will carry that edge into a stronger exit.
That is the gap BluWave closes. One call connects you to an exact-fit AI operator, fast.
