A venture fund built one AI-native platform to replace its doc-sharing, market intel, and CRM stack

The problem

A venture studio raising across multiple vehicles ran its LP relationships on rented software. Documents went out through a doc-sharing service with basic analytics. Market intelligence cost twenty thousand dollars a year. The CRM did not talk to either. Nobody could answer the questions that matter mid-raise: who is actually reading the data room, which LPs are going cold, and who should get a call this week.

What we built

• A data-room platform to replace the doc-sharing service: per-recipient tracked links, watermarking by name and date, one-time-passcode access, and a view-only mode that protects documents on screen. Portfolio updates push straight from the working docs into the platform with no manual upload step.
• Engagement analytics that answer "did they actually read it": section-level attention heatmaps, time on page, and full session playback per LP.
• An engagement scoring engine that rates every LP relationship from 0 to 100 on warmth and deal propensity, with a "going cold" view that surfaces relationships drifting toward silence before they get there.
• A CRM built around how a fund actually works: relationships, shareable custom lists, owner assignment, and bulk import that brought tens of millions of records onto the new stack in about two weeks.
• A morning brief in Slack and email that replaces most of what the market intelligence subscription was for.
• LP onboarding automation wired through the fund's capital-call, KYC, and document-filing systems, so a new LP stops being a week of manual coordination.

Stack