About LP Data
Public private-markets data, assembled with agentic workflows.
LP Data is developed by a single developer with the assistance of coding agents to make fragmented public LP disclosures easier to search, compare, and understand.
The project focuses on public documents that disclose pension fund holdings and performance data. These materials are often published because of local reporting laws, board transparency requirements, FOIA requests, annual reports, investment committee packets, and similar public-records processes.
Behind the scenes, the work involves managing a fleet of coding agents and sub-agents. Some agents find, download, scrape, and parse source documents. Others help normalize the resulting data: reconciling fund and manager names, assigning strategy, sub-strategy, geography, and vintage, and flagging records that need human review.
LP Data is not a replacement for primary diligence or official reporting. It is an attempt to make scattered public information more accessible, consistent, and useful for allocators, researchers, journalists, and other users interested in private markets transparency.
Public sources
Pension disclosures, annual reports, board materials, and other public documents are collected and traced back to source records.
Agent-assisted workflow
Coding agents assist with scraping, parsing, normalization, classification, QA, and data-quality investigations.
Human review
Automated outputs are reviewed, corrected, and improved over time as edge cases and source-document quirks appear.
Contact
For inquiries, data corrections, source suggestions, or feedback on how LP Data could be more useful, please get in touch.
support@lp-data.com