Skip to main content
For funds & portfoliosAnthropic Partner NetworkOfficial OpenAI Partner

One AI program.Every portfolio company.

Train the deal team, then run the same proven rollout across each portfolio company, adapted to its business. One partner, one playbook, and a read-out the operating partners can compare across the portfolio.

A practical 30 minutes: what fits, what to do first, and who does what.

Fund + portcos
One engagement
Repeatable
Rollout playbook
Onsite or remote
Delivery
01

Why run AI at the portfolio level

Operating partners keep getting the same question from every portfolio CEO: what should we be doing with AI? Answering it company by company, with a different vendor each time, is slow and impossible to compare.

More than one of our engagements started the same way. A firm brought us in to train its own team, liked the result, and asked us to do it for the portfolio. The work is different at each company, but the shape is the same: find the few workflows where AI saves real hours, train the people who own them, build what needs building, and measure it.

Running that shape across the portfolio does three things a company-by-company approach cannot. The playbook improves with every company it runs in. The operating team gets comparable read-outs instead of anecdotes. And each new company starts from what worked at the last one, not from zero.

We work with the tools each company already pays for, Claude or ChatGPT, and we keep every company's data separate.

02

What the program covers

Start with the firm, extend to the portfolio. Each piece can run on its own, but they are designed to connect.
01

Deal team and operating partner training

Claude and ChatGPT on the firm's own work: diligence reading, memo drafting, market scans, model review. The team leaves using it daily, and able to judge AI claims in the companies they look at.

02

Portfolio company AI assessment

A short, standard read on each company: where AI would save the most hours, what the team already uses, and what it would take. Written the same way every time, so they compare.

03

Repeatable portfolio rollouts

The same proven rollout at each company, adapted to its business: a leadership cohort, function-specific sessions for finance, sales or operations, and the first builds.

04

Builds where the value is

Agents, automations and connectors scoped per company, and reused across companies when the same problem shows up twice, which it usually does.

05

Portfolio read-out

One view for the operating team: what each company trained, shipped and adopted, and where the next round of effort should go.

06

Data kept per company

Every company's documents, accounts and builds stay in its own environment. Nothing learned at one company's data is carried to another.

Want a sample portfolio plan?

Leave an email and we will send back how we would sequence the first three portfolio companies, and what the firm's own team would get first.

No sequence, no spam. Or book a call directly.

03

Who this is for

Built for firms that want AI to show up in the value-creation plan, not just the pitch deck.

Private equity and growth equity firms with operating partners responsible for value creation
Holding companies and family offices with several operating businesses
Deal teams that want to use AI in diligence and portfolio monitoring themselves
Portfolios where the same functions, like finance, sales and operations, repeat across companies
04

How it runs

Firm first, then one pilot company, then the rest of the portfolio on the playbook the pilot proved.
01Month 1

Firm

Train the deal team and operating partners on their own work. Agree which portfolio company pilots.

02Month 1–2

Pilot company

Assessment, leadership cohort and first builds at one company. The playbook gets written down as it runs.

03Month 3+

Roll out

The next companies run the playbook, adapted to each business, on a schedule the operating team sets.

04Quarterly

Read-out

One comparable view across the portfolio, and a decision on where the next round goes.

05

Who does what

Clear up front, at the firm and at every company.
We do
  • Design and deliver the training at the firm and each company
  • Run the assessments and write the read-outs
  • Build and hand over the agents and automations
  • Keep the playbook current as the models change
You bring
  • An operating partner who sponsors the program
  • A named owner at each portfolio company
  • Time for each cohort to attend and practice
  • Introductions and access at each company
06

Frequently Asked QuestionsWhat teams ask us before they start.

Do we have to commit the whole portfolio up front?
No. Most firms start with their own team and one pilot company. The rest of the portfolio follows once the pilot has shown results the operating team trusts.
How do you handle different tools at different companies?
We work with whatever each company already uses. Some run Claude, some ChatGPT, some both. The playbook is about workflows and people, and it holds regardless of the vendor.
Is portfolio company data kept separate?
Yes. Each company's documents and builds live in its own accounts, and we never use one company's data in work for another. Reusable pieces, such as an agent pattern, are rebuilt per company without the data.
Can you train in person?
Yes. We run onsite sessions at firm and company offices across the US and internationally, remote sessions for distributed teams, or a mix.
What does it cost?
It depends on how many companies are in scope and how much building each needs. We scope it on a call and send a written proposal for the firm and the pilot first.

Talk to us about your portfolio

Thirty minutes with the operating team. Tell us which companies are asking about AI, and we will sketch how the firm and the first pilot would run.