The problem no one talks about in LP reports
Walk through most portfolio companies and you will find capable people doing work that software should handle. They enter data, compile reports, route customer tickets and reconcile invoices. Many of those tasks take an AI system milliseconds, yet they sit with employees earning £40k to £80k per year.
McKinsey estimates that automatable tasks consume 15 to 30% of operating capacity in most businesses. For a company with 200 people, that equals 30 to 60 FTEs of work.
The technology already exists. The constraint is getting working systems into portfolio companies and making sure the teams use them. That constraint is where most AI investment fails: the people who understand the technology and the people who understand value creation are rarely in the same room, and the gap between them swallows the return.
What AI deployment means in a PE context
PE firms are usually offered one of two things: a strategy deck or a deployment.
A deck can help prioritise where AI might be useful, but it does not move EBITDA. Deployment means working inside the portfolio company to:
- Map where teams spend time, which processes repeat and where errors occur
- Build automations that connect the existing CRM, ERP, email and spreadsheets
- Deploy AI agents that can take defined actions rather than only produce text
- Drive adoption through training and process change
Value creation and operating partners should judge progress by what is live and measured. The firms making progress with AI are the ones putting automations into production, not the ones commissioning the most strategy work.
Three high-ROI use cases across PE portfolios
1. Deal origination and pipeline management
Investment teams still spend substantial time screening opportunities by hand. AI can process 10,000 plus opportunities against fund criteria in 24 hours, classify each by sector, geography, revenue profile and deal stage, transcribe deal calls, draft follow-up emails and monitor portfolio companies for news or exceptions.
We have seen origination automation save more than 10 hours per person per month.
2. Finance and reporting automation
Finance teams in many portfolio companies spend 40 to 60% of their time aggregating data and producing reports. An automated workflow can pull and reconcile data across systems, generate management accounts and board packs from templates, flag anomalies and budget variances in real time, and process invoices and payments.
The return combines staff capacity with fewer errors. Automated reconciliation is more accurate than the manual process it replaces.
3. Customer operations
Customer operations is often the largest headcount cost in both B2C and B2B portfolio companies. AI can resolve tier-1 questions such as refunds, order updates and FAQs, or send a complex issue to the right person with the relevant context already filled in. It can also alert customers to delays before they make contact and analyse ticket patterns for product or operating problems.
One portfolio-type deployment helped raise a Trustpilot rating from 4.2 to 4.8 while cutting manual support time by 60%.
The compounding effect
The value increases when the same patterns are reused across a portfolio. After deploying a framework in 5 to 10 portfolio companies, the team carries what it learned into each new build. Work that takes 4 weeks at the first company can take 2 weeks at the fifth, and the playbook improves with every deployment.
That is why we structure PE work as a portfolio-wide partnership rather than a series of isolated projects. The economics improve as the portfolio reuses proven components and operating knowledge.
What to look for in an AI partner
Ask any prospective partner to show you something running in production. They should explain what it does, how it was built and what result it delivers. A roadmap or framework is not evidence of delivery.
The teams that produce results move quickly, work on-site when needed and give adoption the same attention as the build.
Squirrel AI is an AI value creation partner to PE funds and their portfolio companies. We go into the portcos and build, rather than advising from the sidelines, and everything we build is aimed at the P&L. We were investors ourselves before this, so we cover the commercial side as well as the technical.
Nihaar Udathu is Co-Founder of Squirrel AI and previously Head of AI at Healf, the FT's #1 fastest-growing company in Europe. Book a discovery call to discuss AI automation for your portfolio.