My friend, a Partner in a consulting firm, was shocked: “Look, right now I wonder if my job will be around next year”. My friend – let’s call him George – had built his career around CDDs for PE firms. He had just got off a call with a client and vented what happened with me:
“We had done phase 1 on a target for them, and all was set for phase 2. But now they cancelled us. They did 30 expert calls in-house and ran the synthesis with Claude, with the IM and data room”.
“No worries”, I said, “surely it’s just some light-touch work. You’re an expert in this sector – they’ll still want you for the second opinion, and your firm's rubber stamp when going to IC”.
“You don’t understand”, he said, “these megafunds have seen more deals than we have. They have all the IMs, all the comps, and they get data from every banker out there.”
Someone else joined the Teams call, so we switched to our real agenda. But the question stuck with me. Will George’s job still be around next year?
To answer that, let's first look at:
What do CDD consultants do all day?
PPT slides were still 4:3 (print size) back when I was at McKinsey, but the deliverables of a CDD remain the same:
)
Let’s look at each of these 7 points and their likelihood of being automated with AI.
Use AI to analyze all data. Pretty straightforward. Claude shreds through the data room faster and more diligently than your associate Clark does. You’ll have to double-check both of them, though.
Sourcing data is as hard as you make it, and effort generally pays off. Data that is available to all is alpha for noone. Finding the right experts, asking the right questions, seeing patterns across a mosaic of sources, telling signal from noise and knowing when you have enough data for conviction - all set apart the good investor/consultant from the mediocre.
Crafting an independent perspective is not as easy. Some CDDs might feel like a pure put-up job, without room to question the narrative. But any consultant worth their salt will have a perspective of their own. If you don’t have one, you’re probably not senior enough. That said, I’d think of LLMs as any writing assistant. You can’t let it think for you, but once you have a draft of your own, use it for feedback and inspiration.
Flag commercial risks. Same as the above – you’d be dumb not to use AI for additional inspiration (we turn all stones, remember?), but you’d be equally dumb to use it without first thinking for yourself.
Banker comms. Sure, man, here you go: 8-10%.
Rubber stamp to IC. This one is hard to replace. It might work in firms investing their own money only, but PE today is institutionalized and there is a rulebook. If you don’t tick the boxes, many institutional hands are tied. Now, this is also the least exciting thing for a strategy consultant to be. But your firm’s brand paired with your independent perspective – that’s your moat.
Storytelling. Claude always tells me what I want to hear.
Returning to our graphic, it looks like this:
)
Ultimately, PE investors are smart and competitive. The payoff for making great investments far outweighs the $500k you might pay a consulting firm.
From what I'm seeing, tech tools (whether AI or not) allow you to screen more deals than you would do manually. They also let you drill deeper into every case you double down on. That's all good - but it rather widens the top of your funnel than replacing downstream steps.
)
So what’s the verdict - will AI replace consulting CDDs?
George – you won’t lose your job to AI. But you’ll lose your job to a consultant using AI. All the big consultancies are rushing the same way: build tech capabilities to deliver better work.
For CDD-focused teams, this boils down to mainly three things:
1. Accumulate data
Sanitize and share past project materials.
Gather transcripts of all expert calls, client meetings, steercos. Consultants transcribe thousands of calls every week on Inex One. They analyze them in our embedded AI tools – or in their own, through our MCP server.
Figuring out data access rights is central. Should everyone see all content? What if two teams serve competing clients? At Inex One, we’ve solved access rights. You can see whether your colleagues have covered a topic, and request access to individual transcripts and projects. This works both in-app and through MCP.
)
2. Upskill the team: in productivity tools and content.
Specializing your team (or entire firm) early on helps. Sector-specialized consulting firms are outpacing incumbent generalists lately. Gone are the days when consultants just helped “read your watch”; now – more than ever – you’ll need to bring expertise.
Hiring batches of smart youngsters, skewing towards engineers, helps with tool adoption. You need critical thinking, and integrity to question and challenge the data (i.e. managing AI) rather than taking it as given (simply using AI).
3. Eliminate donkey work
Freeing up time for thoughtful analysis. It’s a wide bucket. At Inex One, we help specifically with:
Simplifying expert network management. All expert networks in one platform. You cast a wide net, see all available experts, and eliminate the email chaos.
All call transcripts in one place, analyzed with AI.
)
Punters have predicted the disruption of consulting many times before. The pyramid might change, with fewer juniors per senior partner/expert, but I don’t see AI replacing you entirely – if you use the tools available and keep an independent mind. There will be a phase 2, George.
That's a wrap. Let me know what you think I missed, or check out Inex One here.
Want to read more? Shoutout to five people whose writing influenced me:
Christoph Nicau of Valantic writes about how fake market data infects LLM training sets, and the circular “death loop” this creates.
James Agres of 2nd St Strategy shares his experience building AI tools for CDDs – where AI works and where it doesn’t.
James O’Dowd writes about the competitive pressure from AI labs on traditional (mainly ops-focused) consulting firms, and the opportunity for sector specialized firms.
Beltrán Simó describing what in-house AI tools at large consulting firms do, and what skills will be relatively more & less valuable going forward.
Dave Demasi shares learnings from building one of those AI tools - incl. pivoting back from “AI-only” to “AI+human for judgment”.
