Built a complete AI agent system for Watershed Ventures that handles nearly every part of a VC partner's daily workflow. The system automates research, company and people evaluation, investment memo drafting, deal flow tracking, and monitoring of early-stage and stealth company developments as they happen.
What It Does
Research + Evaluation
Agents independently source, enrich, and score companies and founders across data providers, accelerators, and early-stage funds. Cross-references funding history, headcount trends, LinkedIn profiles, and more to build a full picture fast.
Memo Drafting
Generates structured investment memos from enriched data, saving partners hours of manual writing while keeping the output grounded in real data points.
Stealth + Early-Stage Tracking
Monitors rapid developments and changes at stealth and pre-seed companies that traditional tools miss. Surfaces signals before they hit the mainstream deal flow.
Daily Partner Workflows
Handles CSV imports, database updates, portfolio monitoring, and recurring data hygiene tasks. Partners interact through an internal platform to view results, filter companies and people, and manually adjust when needed.
Agent Architecture
Specialized agents for company retrieval, data enrichment, memo generation, and database management, all coordinated through a central orchestrator. LLM-driven reasoning (Google Gemini) lets agents decide when to query APIs, update records, or flag incomplete information.
Impact
- • Cut manual research and data entry time by over 50%
- • Always up-to-date view of emerging startups, stealth companies, and fund activity
- • Foundation for autonomous deal-sourcing workflows that keep getting better over time
- • Internal platform used daily by the team to interact with the system
- CompanyWatershed Ventures
- RoleAI Software Engineering Intern
- DurationMay 2025 - Present
- StackPython, Google Gemini, Supabase/Postgres, AWS, Next.js
- Key TechnologiesMulti-Agent AI, Database Design, API Integration, LLM Reasoning