Private Equity × AI Adoption
Protect Your Investment.
AI-Enable Your Portfolio.
SJT Digital helps PE firms work out where AI actually creates value across their portfolios, then delivers it: at fund level, and inside the companies themselves.
You run the fund
Which of your companies' AI claims stand up, where the value actually is across the portfolio, and how to get it delivered without a programme.
What this looks like for funds ↓You run one of the companies
Your investor has asked what you are doing about AI. Where it genuinely earns its place in everyday work, and how to get the first thing working.
For portfolio companies →The Shift
Product and engineering economics have changed.
Most PE portfolios haven't caught up.
AI is changing both sides of the equation. What gets built - discovery and validation cycles that took months can now be compressed into weeks. How fast it ships - work that previously required coordinated teams, sprint planning, and weeks of handoffs is increasingly achievable by a small team moving with the right tools. The cost of building software is falling, and it will not stop.
Private equity firms are beginning to ask the right questions: why are our portfolio companies still running six-month discovery cycles and the same engineering headcount they had three years ago?
AI doesn't eliminate the need for engineers; it changes how many you need, and what the best ones are worth. Done well, this is an opportunity to right-size engineering cost ahead of exit: a leaner, faster team that delivers more. Done poorly, or not at all, and you're explaining to your LPs why your engineering burn looks identical to three years ago while your competitors have already made the shift.
The playbook for this was built inside PE-backed software businesses, with real teams and real results. It transfers.
Where it usually starts with a fund
Every company in the portfolio reports on AI.
Almost none of it has been checked.
By now most funds have asked their companies where they stand on AI: how exposed the business is, how far along they are, what they are doing about it. The answers feed board packs, valuation discussions and increasingly the auditors. And they are written by the people being measured on them, in a company that has read the question its own way and has every reason to give a reassuring answer.
Nobody in a fund is well placed to argue with that. The investment team knows the businesses but not the technology; the operating partners are stretched across the portfolio; the advisers who could judge it are the ones who would then sell the fix. So the self-assessment stands, and the fund carries a view of its AI exposure that it has never tested.
SJT Digital tests it. Take what the companies have already said, hold every one of them to the same written standard, and give the investment team a view they can defend: which positions are sound, which need a proper conversation with management, and which cannot be relied on yet. The standard stays with the fund afterwards, so the next cycle can be run in-house.
Inside the companies
The portfolio view tells you where to spend.
This is what spending it looks like.
Four engagements SJT Digital runs inside individual portfolio companies, each described in outline. Most funds start with the first or the last.
Make delivery cost less
Resetting what a programme costs and how long it takes
A small group of expert engineers with AI tooling, insulated from governance layers and approval queues, delivers what a conventional programme takes multiples longer to produce. The interesting part is never the speed. It is what the speed reveals about how much of the original estimate was structure rather than work.
A core product rebuild estimated at 18 months, delivered in 90 days, on time and to strong customer reception.
Build it, don't scope it
Working software instead of a delivery queue
Some problems do not need a roadmap, a business case and a place in the queue. They need someone to build the thing. Simon still writes code, and a good deal of what SJT Digital delivers is software rather than advice: production applications with authentication, access control, reporting and an audit trail, on the company's own infrastructure.
Spreadsheets to a deployed, enterprise-grade application in one week, for the cost of a monthly software subscription.
Make the programme legible
Governance an investor and an engineer both believe
Large programmes go dark long before they go wrong. Status becomes a negotiation, forecasts get re-argued every cycle, and nobody can say what is actually true. SJT Digital puts in the reporting line that rolls delivery up from the work itself, so the board sees the same picture the engineers do, and decisions stop being re-litigated.
One view of a multi-workstream programme, produced from the delivery data rather than assembled by hand each month.
Get the companies moving
AI in everyday work, one company at a time
Most portfolio companies are past the licences-and-a-pilot stage and stuck there, and their CEOs are asking the fund what to do next. SJT Digital goes into a company, separates the few places where AI genuinely earns its keep from the many where the fix is cheaper, and gets one thing working with the people who will own it.
Runs as a portfolio programme: a webinar or CEO forum first, then company by company. The page your companies can read →
Engagements are described in outline only. Client names, sectors, portfolio composition, budgets and any figure that would identify a business or its financial position are withheld, including where no confidentiality obligation requires it. Detail is available in conversation, within whatever the relevant agreement allows.
Background
Simon Taylor
Simon is a fractional CTPO with over 20 years in product and engineering leadership across PE-backed businesses, scale-ups, and enterprise. He has held the line between investor expectations and engineering reality, and knows what both sides need to hear. He is also still building. He works in Claude Code every day, and a good deal of what he delivers is software rather than advice.
Most recently inside a large PE-backed enterprise software business, he ran two AI delivery experiments that have since become the reference points used with every PE client. One proved that a small team with the right tools can compress an 18-month product rebuild into 90 days. The other proved that a single developer, given AI tooling and no bureaucratic overhead, can deliver production software in a week for the cost of a monthly subscription. Since then he has worked at fund level, validating AI claims across a whole portfolio, and inside portfolio companies as fractional CTO on programme governance and delivery.
Both projects pointed to the same pattern: in most organisations, capability is rarely the bottleneck. Structure, process, and accumulated organisational friction usually are. The job is to assess where a company actually sits, build a realistic roadmap, reduce the friction, and construct the investor narrative around what becomes possible once that is done.
He works at both levels. Strategically, discovery and roadmapping: assessing AI maturity, identifying what's realistic and what's premature, and framing the business case. Practically, sitting with the engineers, writing code alongside them, and showing rather than telling. The two reinforce each other in ways that purely advisory work rarely achieves.
Previously: product and engineering leadership at LearnPro Group, Filtered, and a range of PE-backed and VC-backed software businesses across the UK and Europe.
Let's Talk
If you're a PE firm looking to accelerate AI adoption across your portfolio.
A short call is the right first step. No deck, no proposal: just an honest conversation about whether there's a fit. If you run one of the companies rather than the fund, start here instead.