We Trust AI Agents That Break Every Day - Shy Ruparel, Temporal
62% of AI agent teams lose significant time or revenue due to reliability issues
“deciding that AI is what we do doesn't necessarily mean successfully implementing it”
Temporal's Shy Ruparel presents findings from a survey of 554 experienced engineers actively using AI agents, identifying four major gaps between agent adoption and production success. Despite 94% of fintechs calling AI core to their business and KPMG projecting $3 trillion in productivity gains by 2025, 62% of teams report significant time or revenue losses from reliability failures. The talk frames long-running, multi-system enterprise workflows—loan processing, KYC, fraud screening—as particularly vulnerable when AI agents are inserted mid-process.