Loop Engineering and the Future of VC in the Age of AI

Loop Engineering and the Future of VC in the Age of AI How AI-Native Funds Source, Evaluate, and Remember Presented by BFC AI Academy — with Chinat Yu AI is changing venture capital—but the biggest shift isn’t better prompts. It’s better loops. The funds pulling ahead are beginning to build repeatable AI-powered workflows that continuously source companies, evaluate opportunities, capture institutional knowledge, and improve over time. A prompt answers once. A loop compounds. Join Chinat Yu for a practical, one-hour BFC AI Academy session exploring what an AI-native venture fund actually looks like—and how investors can start building these systems today. No technical background required. What We’ll Cover 1. How Do You Extend Your Network? Venture deal flow has always been network-bound: you see what your network sees. AI agents are changing that equation. We’ll explore how AI-native funds can build sourcing loops across: Demo days and startup launchesFounder communitiesPublic signals and emerging companiesWarm-introduction networksContinuous opportunity discovery The goal: expand your fund’s reach without expanding partner hours. 2. How Do You Build an AI-Native Evaluation Platform? What if the first-pass investment memo took 20 minutes instead of three days? See a live demonstration of an AI evaluation agent that can: Read sources → evaluate a company against an investment thesis → score key dimensions → draft a first-pass investment memo while keeping human judgment exactly where it matters most. This isn’t about replacing investors. It’s about giving investors more leverage. 3. How Do You Build a Fund That Remembers? A venture fund’s most valuable asset isn’t just its portfolio. It’s everything the team has learned: Every pitch.Every pass.Every founder conversation.Every follow-up.Every investment decision. Yet much of that knowledge disappears into inboxes, documents, and individual memories. We’ll explore what a true “company brain” for a venture fund can look like—using shared agent memory, connectors into existing tools, and AI agents that continuously keep institutional knowledge current. Inside YC’s QM Y Combinator recently open-sourced QM (Quartermaster), the multiplayer AI-agent harness YC uses across its own operations. We’ll break down: What YC actually open-sourcedHow an agent-based operating system is structuredWhat QM can—and cannot—doWhat it takes to deploy a similar systemWhere a VC fund should start first What You’ll Walk Away With By the end of this session, you’ll understand: The Loop Engineering FrameworkHow to turn repetitive fund workflows into continuously improving agent loops. The Three Loops of an AI-Native FundSourcing, evaluation, and institutional memory. A Live AI Deal-Evaluation WorkflowSee an agent research, score, and draft a first-pass investment memo from end to end. The Anatomy of YC’s QMUnderstand the architecture behind YC’s newly open-sourced agent system. Three Things You Can Build This WeekPractical workflows investors can begin implementing with tools like Claude Code—without building a full engineering team. Your AI-Native Fund DiagnosticIdentify which workflow your fund should automate first and where human judgment should remain. Who Should Attend? This session is designed for: VC Partners & Principals exploring AI-native fund operationsAngel Investors & Syndicate Leads looking to dramatically increase leverageFounders who want to understand how AI-native investors may evaluate startupsOperators moving into venture capitalAnyone interested in the future of AI-powered investment workflows No technical background is required. This is a lecture with live demos + Q&A, not a coding build-along. Workshop Format 60-Minute Live Session ~40 min: Lecture + live demonstrations~20 min: Live Q&A About the Instructor Chinat Yu Chinat brings elite credentials and practical experience: Stanford Learning, Design & Technology graduate with exper…