Agentic Search: Rethinking Search Away From “Top 10 Results”

Saturday, April 18, 2026 · 11:00 AM – 12:30 PM · Classroom 205

Tireless agents can search on behalf of users. Especially in domains where no stone can be left unturned: recruiting, legal research, and more.

Yet these agent often search errantly. A furniture store user wants a “bistro table” - to a consumer, that’s a small outdoors table. Yet the naive LLM assumes the user must want restaurant supplies. How do we ensure our fairly dumb AI intern gets supervised, steered in the right direction

Out of the box that just adding reasoning to a rather dumb search improves search quality. But we can integrate a judge or reranker to push / pull the naive agent towards what its supposed to be finding. We can go beyond this, breaking our assumptions of what search is towards comprehensive crawls of legal discovery corpuses and weeks-long exhaustive searches to find every possible candidate fro a job req.

About the Speaker

Doug Turnbull

Retrieval Engineer

In 2012, Doug got bit by the search bug and he's still trying to keep up. From full-text search, to Learning to Rank models, to search agents that generate their own code, he knows the overwhelming landscape first hand. Yet Doug still works to deeply understand the what / how / why, and help teams use these technologies practically, distinguishing hype from reality. He’s led search at Reddit, Shopify, and Wikipedia, authored Relevant Search and AI Powered Search, and advised 100+ organizations over the years - all in pursuit of the same question: how does search actually work?