AI readiness across your portfolio
Where each company actually stands on data and AI, which use cases are worth funding, and what it takes to get them used. Delivered by one person, working directly with the management team.
Three stages
In this order, because a use case identified in a company whose data cannot support it is a write-off. Each stage stands alone if that is all you need.
Current state and readiness
What data the company holds, what state it is in, and whether it can support what is being proposed. Alongside the two things that usually decide whether AI lands at all: whether staff have the skills, and whether anyone has set a position on governance.
- Data: what exists, where it sits, quality and accessibility
- People: skills, appetite, and what staff are already doing with public AI tools
- Governance: policy, risk position, and what the board can currently defend
- A maturity score across eight pillars, repeatable so you can re-run it in twelve months
- A go, no-go or not-yet on any major investment being proposed
Use case identification and prioritisation
Working through the business function by function to find where AI would change a number, then scoring what is found against the data the company actually has.
- Opportunities scored on impact against feasibility
- Each one sized against the data available to support it
- A shortlist worth funding, and the ones to leave
- What each would cost to build and to run
Embedding
The part that decides whether any of it reaches the numbers. Most portfolio companies do not need a bigger idea; they need the first one used.
- A governance framework the board can sign off
- Training pitched separately at leadership and at staff
- One first use case built and measured
- A twelve-month roadmap with decision points
What I look for
The same issues come up across portfolio companies.
- Manual work sitting in the cost base. People retyping data between systems, reconciling by hand, or rebuilding the same report every month.
- Reporting nobody trusts. Two teams producing different numbers for the same thing, so decisions get made on instinct anyway.
- Key-person risk. One person who understands the data, with nothing written down.
- AI in use without governance. Staff already putting company and customer information into public tools, with no policy and no record.
- Platform spend without adoption. A tool bought, partly implemented, and quietly worked around.
How I approach it
Readiness comes before opportunity. Establishing what is true first disqualifies ideas early, which is cheaper than discovering the same thing six months into a build.
Sometimes the answer is that nothing needs building.
More on AI enablement and governance · More on data strategy
Looking at a portfolio company?
A confidential conversation, under NDA if you prefer. Tell me the decision you are trying to make and by when.
Book a call