AI value creation across the portfolio
Two places AI earns its keep in a fund: the fund’s own operations, and value creation inside the portfolio. This page is our thesis on the second.
Email us
Why AI value creation stalls in portfolio companies
A fund theme is not a project
Nobody inside the company owns a theme. Without a named owner and a budget line, the theme survives two board meetings and then quietly stops.
Pilot KPIs that never reach the P&L
A pilot metric the finance function does not recognise cannot become an EBITDA effect, however good it looks in the deck.
No engineering capacity
The company already has a roadmap and it is full. An AI programme that needs the same engineers competes with the product.
From fund thesis to EBITDA effect
Cost, price and multiple
Three levers, and only one or two are available in a given company. Picking the wrong one costs a year.
A 100-day pattern
Scope, build, measure, hand over, in that order, so the quarter ends with a running system rather than a plan for one.
What we do in a portfolio company
We do the build ourselves, alongside the team that will keep it after we leave, so no capacity is taken from the product roadmap.
AI in the fund’s own operations
Sourcing, screening and diligence support are the same kind of work as portfolio automation, one level up.
Read the long version
AI value creation in portfolio companies
The long-form argument: the three levers, the 100-day pattern, and the measurement path from a pilot KPI to a line in the P&L.
Where this comes from
This page is a thesis, not a case study. Our published project work sits in medtech and B2B technology, and the pattern is drawn from it.
Related
AI value creation in portfolio companies
The full article, with the measurement path from pilot KPI to EBITDA.
Read onAI strategy and agents
The service that does most of the building inside a portfolio company.
Read onB2B tech and SaaS
The commercial automation most portfolio companies start with.
Read onQuestions
Do you sell to funds or to portfolio CEOs?+−
Both, and the work differs. A fund buys a pattern it can apply across companies. A portfolio CEO buys one system that has to run by the end of the quarter.
How long before something is live?+−
The 100-day pattern targets a working system inside one quarter, with the measurement agreed before the build starts.
What does the fund get?+−
A measurement path from the pilot KPI to a line the finance function recognises, and a system the company can keep running without us.
Do you work across several companies at once?+−
Sequentially rather than in parallel. We are a small team and we do the build ourselves, which sets the pace.
Still have questions? Get in touch