Engineering Agency out of the Happy Path — Matthew Jewkes, Standard Cybernetics
Source
youtube.com
Author
AI Engineer
Date
Why it matters
Shows how splitting model errors, missing organizational answers and genuinely new cases into separate routes lets a routine path reach 99.95% accuracy while escalating the rest, a reusable pattern for high-reliability agent systems.
Key takeaways · AI-distilled
At Clarivate, Matthew Jewkes's team indexed drug development documents arriving at tens of thousands of new tasks a month, and an early AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → approach reached only around 80% to 83% accuracy.
Jewkes says the routine path became a tightly specified automation, with uncertain and unusual cases escalated instead of pushed through.
Escalated cases were grouped for experts into three piles: critical exceptions to clear, ambiguous low-value work to freeze, and unusual demand that might justify a new product.
One expert used that structured backlog of exceptions to form an investment thesis before a major customer raised the same need, according to Jewkes.
His bet is that companies will gain more from managing collective expert attentionThe mechanism that lets a model weigh which earlier words matter for the word it's currently processing — the core operation of a transformer.Full definition → and judgment than from choosing a particular orchestration framework.
Terms in this piece · Glossary
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
attention — The mechanism that lets a model weigh which earlier words matter for the word it's currently processing — the core operation of a transformer.
videoYour KV Cache Can Spot Attacks on Your AI Agents - Ari Azarafrooz, Intrinsec AI
videoWill AI Agents Pretend to Be Human to Get Past a Check? - Maty Bohacek, Stanford
videoThe One Thing Every AI Agent Framework Agrees On - Mislav Ivanda, Daytona
videoAI Writes Code Faster Than Uber Can Merge It - Preetam Dwivedi & Manjari Akella, Uber
videoTrading Desks to Clinical Trials: Parallels in Applied Vertical AI — Ayush Bhardwaj, Allos AIAI Engineer
videoWhy We Killed Our Multi-Agent Pipeline — Subbiah Sethuraman and Abhilash Asokan, ZS AssociatesAI Engineer
videoFrom Ambient Documentation to Clinical Intelligence — Chaitanya Asawa, AbridgeAI Engineer