{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"The First Bet","title":"EP7: Brian Bell - AI as decision support in Venture","html":"","width":"100%","height":180,"duration":2182,"description":"Brian Bell on Building a Venture Investing System for Low-Information Decisions\nMartin Tobias talks with Brian Bell, Managing Partner of Ignite Ventures, about how he makes startup investment decisions when there’s little information and a lot of pressure. Brian shares how years inside AWS and Microsoft shaped his eye for talent, product quality, and market timing, and how he’s now combining pattern recognition with AI to underwrite early-stage companies faster and more consistently.We discuss Brian’s framework for sourcing through YC, scoring founders and startups with a weighted model, and using AI as a thought partner rather than a replacement for judgment. The conversation also covers pivots, fragility, red and yellow flags, and how to learn from both wins and misses over time.\nKey topics\nBrian explains why he bootstrapped deal flow through YC, where 20,000 applications are filtered down to about 150 to 200 startups per batch, creating a high-quality pool for fast decisions.\nHe describes why he raised a fund after running syndicates, mainly to move quickly when rounds closed early, valuations changed, or founders didn’t want to syndicate broadly.\nBrian says the strongest early signal is still founder quality, including star power, recruiting ability, coachability, and velocity of learning.\nHe and Martin discuss how timing matters in venture, and how a product can be too early, on time, or too late.\nBrian shares that his team built an AI-assisted scorecard using about 20 features, trained on thousands of past calls, pitch decks, resumes, and YC outcome data.\nThe model outputs a rank from one to five, plus separate scores for power law potential, fragility, and red and yellow flags.\nHe says the AI helps stack rank YC batches and pre-sort the best opportunities, but he still manually reviews everything and often adjusts feature scores based on context.\nBrian highlights key fragility vectors like founder fragility, market fragility, product fragility, capability...","thumbnail_url":"https://img.transistorcdn.com/svm9JgXvtaknzsrbiGn4yO8TiDK8Ptj7rMt6cBbfV7o/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS83MmVm/YTg2NTQyMmM0YzMz/ZGNjMTg3ZGE1OGM2/NWQzNy5wbmc.webp","thumbnail_width":300,"thumbnail_height":300}