All workFILE 04 / 16

Private markets

Reading private-markets statements at scale

Shipped for A private-markets data platform · led by Nikhil Bery

01

The problem

Fund performance lived in 150+ inconsistent PDF statements from major VC and PE firms, and extracting fund metrics by hand did not scale.

02

The constraint

The statements arrived in 150+ inconsistent PDF layouts, so the numbers had to be read reliably at scale, not keyed in by hand.

03

The system

A multi-engine OCR and parsing pipeline for hierarchical fund tables (IRR, MOIC, DPI, TVPI), with QA dashboards and CI/CD.

FIG. 01 · SYSTEM FLOW · 5 NODES · 4 FLOWS

150+ fund PDF statements → multi-engine OCR. multi-engine OCR → hierarchical table parser · IRR·MOIC·DPI·TVPI. hierarchical table parser · IRR·MOIC·DPI·TVPI → extracted fund metrics. hierarchical table parser · IRR·MOIC·DPI·TVPI → QA dashboards

04

The outcome

150+ financial statements from 6+ major VC and PE firms processed, with fund metrics extracted automatically instead of by hand.

← FILE 03 · A WEALTH-MANAGEMENT FINTECHMillions of securities, one grounded answerFILE 05 · TWO NORTH AMERICAN PRIVATE-EQUITY FIRMS, ACROSS $10M-$50M PORTFOLIO COMPANIES →A data spine for mid-market portfolios
Start your own file