How a US$120M early-stage fund rebuilt its deal pipeline with AI — and stopped losing rounds to faster funds.
Vertex had grown to 340 deal decks/month. Their one associate couldn't screen them all — partners suspected they were missing gems. Data-room diligence for a competitive round took 11 days on average; twice they had lost deals to funds that moved faster. IC memos were rewritten from scratch every time. LP reporting consumed a full week per quarter.
A deal-screening co-pilot ranked every incoming deck against Vertex's explicit thesis (fintech-adjacent, MENA/SEA, revenue >AED 400k ARR, founder-market fit). A private data-room AI read entire data rooms and produced a 6-page brief with red-flag callouts. A memo assistant drafted structured IC memos from associate notes and market pull-ins. All models ran on Vertex's controlled infrastructure — no data-room content ever left their environment.
"Our deal-sourcing pipeline used to be gut feel. Now we screen 40× more opportunities against thesis — and back the right ones faster than the market."
Investment: AED 380k across 8 months. Return: estimated 3.4% IRR uplift on a US$120M fund equates to material multi-million-dirham LP value. Net ROI: qualitative but partners describe it as 'unquantifiably valuable.'
Get in touch and we'll map what will (and won't) work — with ROI numbers in AED.
