Sourced examples from real firms, sorted by the job they do. Every card shows where the claim came from and who made it. Every use case was reviewed by an editor before it went up; vendor claim means the source is the vendor itself.
Guinness Ventures' head of technology describes using AI for filtering, research, benchmarking and drafting, with every output checked and investment committee decisions kept human because the models are 'frequently wrong'.
Hanover Park publishes a downloadable skill: ask Claude to 'Draft our Q2 LP letter' and it pulls performance, capital activity and portfolio developments through a read-only connector, adding a tie-out appendix.
Primary's job post sketches the emerging role: one leader owning AI systems for sourcing, diligence, portfolio support and firm operations.
Marc Palet shares the stack he uses for knowledge-base upkeep, research and reporting, and why recurring operational tasks now go to agents rather than a junior hire.
AngelList's MCP connector gives GPs read-only access to their fund records from Claude, ChatGPT and other MCP tools, so questions like 'what's my remaining reserve' get answered in chat.
Five KV talent partners support searches across hundreds of portfolio companies at once.
A roundup including Correlation Ventures, whose predictive model trained on 100,000+ financings lets it make co-investment decisions in about 48 hours.
Serac built a screener that matched inbound deals to its thesis, auto-booked meetings and auto-passed the rest.
Alongside its Seed 100 list, Business Insider collected how leading seed investors use AI for sourcing, prep and diligence.
Announcing its 2026 analyst class, USV explains how six months of deploying an agent fleet changed the role: less note-taking and lookup work, more time meeting founders and building conviction.
Krishin Parikh walks through a Claude Code setup for diligence: market and competitor research, founder background checks and a structured write-up, with each claim tied to a source the investor can verify.
Point Nine managing partner Christoph Janz built a firm-wide knowledge hub in 13 days without writing code himself.
Airtree's research agent produces a company brief in about three minutes, work that used to take an analyst hours.
Korean VC TheVentures built Vicky, an AI analyst that reads decks, founder profiles and transcripts and returns a Yes/Maybe/No rating with reasoning.
Fika Ventures open-sourced a small tool that checks portfolio and target companies' career pages on a schedule and posts to Slack when a new role appears, turning hiring into a live momentum signal.
HP Tech Ventures runs a team of specialized agents on Claude Code and Obsidian, including a Chief of Staff orchestrator.
A cautionary tale: a Quiet Capital partner spent about 50 hours building a personal command center with Claude Code across email, calendar, notes and CRM.
USV says it roughly doubled its team by adding AI agents instead of hiring.
USV gave each internal agent a name and a job: Sally writes up meetings from Granola transcripts, Ellie watches team email, Felix handles finance data, Arthur supports deal analysis and Connor tracks calendars.
USV is onboarding Leo, a legal counsel agent that reviews legal documents alongside the team, the newest member of its named agent roster.
USV's Ellie monitors the team's group email threads and turns mentions of companies and people into structured records in the firm's internal database.
Canonical's two partners build their own agents for thesis work, diligence and founder support, arguing nobody knows their workflow better and investors in AI should 'feel the curve'.
SwiftLaw generates LPAs, PPMs and subscription documents from a fund's term sheet, a job usually billed by the hour by fund counsel.
Younger VCs describe using ChatGPT and Google's NotebookLM to find deals, prep for pitches and get up to speed on new markets quickly.
AI meeting notes.
AI-native CRM.
Cap tables and fund administration.
Startup discovery and sourcing.
Relationship intelligence CRM.
Fund infrastructure.
