[ENGLISH SESSION] This session provides a data-driven framework for evaluating AI/ML investments in startups. It walks through the AI project lifecycle — from ideation to scale — highlighting the "Valley of Death" where 87% of AI projects fail between Proof of Concept and production. Attendees will learn the four main failure modes of AI projects, a practical risk matrix (technical vs. business risk), five early warning signals that an AI initiative is in trouble, and a concrete Fund / Scale / Kill stage-gate framework for making portfolio decisions. The session concludes with a 10-question due diligence checklist that investors can immediately apply to score AI startups. Ideal for business leaders and investors navigating the startup AI landscape.