Verification Gap: What Separates LLM Demos from Prod Agents | Andriy Batutin | LLMday Warsaw 2026 Q1 скачать в хорошем качестве

Verification Gap: What Separates LLM Demos from Prod Agents | Andriy Batutin | LLMday Warsaw 2026 Q1

LLMday Warsaw 2026 Q1 - February 12 Grab your ticket for the next LLMday: https://www.llmday.com Upcoming LLMday CFPs: https://cfp.ninja/?q=llmday&status=op... Chapters 00:00 Intro & Switching to English 00:17 What Is “Verifier Primacy”? + Speaker Background 00:54 The Coming AI Failure Wave: Stats, Benchmarks & Why It Matters 02:22 The AI Adoption Paradox: Move Fast vs. Career Risk 03:02 Why AI Is Hard: Next-Token Chaos, RAG Limits & Agent Explosion 05:01 Insurance as a Case Study: Data Silos, Tribal Knowledge & Risk Levels 06:39 Human-in-the-Loop Doesn’t Scale: The Rubber-Stamp Trap 06:59 Claim Assistant Horror Story: How a Bad Payout Happens 09:54 Stop Using Humans as the Last Line of Defense 10:23 Evaluation Methodology: Error Analysis → Taxonomy → Evaluators → Metrics 11:15 Better UX for Verification: Show Sources, Retrieval & Reasoning 12:06 Building the Verification Stack (6 Layers) + Escalation Paths 13:32 Operationalizing Verification: From “AI Firewall Engineers” to Business Moat 14:25 Wrap-Up & Contact 15:02 Q&A: Can We Get Deterministic Guarantees? 16:01 Q&A: Verification vs. User Experience in Consumer Products 18:22 Q&A: High-Stakes Domains, Checkpoints & PDF-to-Rules Challenges 20:44 Q&A: Adversarial Users, Prompt Injection & Realistic Limits 21:27 Final Thanks & Closing