Free Lesson

Why your hybrid search sucks

Hosted by Doug Turnbull (Maven)

Wed, Aug 12, 2026

4:00 PM UTC (1 hour)

Virtual (Zoom)

Free to join

Invite your network

Go deeper with a course

Build your own vector database
Doug Turnbull
View syllabus

What you'll learn

It's not enough to merge dense + sparse results

Why naive RRF and interleaving doesn't solve hybrid search's chicken and egg problem

The missing intent layer for selecting retrieval strategies

Retrieval strategies (dense, sparse, hybrid) depend on understanding the user's intent to route to the best approach

Why query understanding + metadata filtering become crucial

Why techniques like late interaction push the search world back towards sparse vector techniques

Why this topic matters

Hybrid search isn't merging different systems. It's a single search engine. Today's systems combine dense + lexical retrieval to get the best of both worlds. Just taking top 1000 vector + top 1000 lexical isn't enough. You'll need to filter the vector results by almost everything in the lexical side. With today's search systems, think of this as one system, not two,

You'll learn from

Doug Turnbull (Maven)

Led teams at Shopify, Reddit, Wikipedia

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 endless landscape first hand. Yet Doug wants 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?

Previously at

Shopify.com
Reddit
LexisNexis
OpenSource Connections
See all products from Doug

Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.