Receipts for AI: Trust You Can Prove

AI is moving from chat windows into real engineering workflows: suggesting code, preparing changes, triaging incidents, and helping teams make operational decisions. But as automation gets closer to production, a harder question appears: how do we prove what the AI saw, what it decided, what validated it, and why it was allowed to proceed? This session explores the idea of decision receipts for AI-assisted engineering: a practical evidence layer around agents, automation, and operational workflows. Instead of trusting AI because it sounds confident, we can make its actions explainable, reviewable, and traceable. For engineers, founders, product leaders, and anyone building AI into real systems, this is a conversation about moving from "AI did something" to "we can prove what happened." Come for the receipts. Stay for the conversation.