An agent that solves a task four times out of five is impressive in a demo and unshippable in production: state-of-the-art agents scoring 80 to 95 percent on isolated steps drop to 20 to 40 percent on long-horizon tasks. This talk distills a survey of roughly 85 papers on agent reliability into a builder's toolkit: a two-axis taxonomy of where errors are born and how they spread, the self-correction paradox (the model checking the work is the model that made the mistake), and the inference-time "decode tax" that makes reflection-style fixes cost more than they return. You'll leave with a vocabulary for diagnosing your own agent failures, a map of which mitigations target which failure types, a public dataset of 50 annotated failure traces, and an honest account of what nobody has solved yet.
Sujoy Datta Choudhury is a Software Architect with 20+ years of experience in distributed systems, AI platforms, and cloud infrastructure. As Technical Architect at Amazon Ads, he oversees platforms processing billions of daily requests and $1.1B+ in ad volume. Previously VP at BlackRock, he built mission-critical Fintech platforms and founded their global AI hackathon. An AWS Labs maintainer and published researcher, he holds an M.S. in Data Science.
