.png&w=1536&q=75)
Free Lesson
Optimize Structured Data Retrieval With Evals
Part of The AI Evaluation Handbook
45 min
May 8, 2025 8:00 PM
Virtual (Zoom)
In this video
What you'll learn
Identify blind spots in your current evals.
Understand how hybrid search, boosts, filters and reranks on structured fields shape your results in unexpected ways.
Build search evals that account for semi-structured data
Most real world data has structure. Learn how to properly account for this when writing evals.
Build more realistic evaluation datasets
Learn practical techniques for creating on enhancing datasets with web scraping and LLM judges in clever ways.
Why this topic matters
It’s time to stop evaluating on queries like “hotel nyc” while your users search “top family friendly hotel near manhattan midtown under 200”! Daniel will share his experience evaluating retrieval quality across multiple domains (travel, jobs, fashion). Take your evals beyond text-on-text matching and close the gap between your evals and production.
You'll learn from
Daniel Svonava
CEO at Superlinked. Previously ML TL at YouTube.
Hamel Husain
ML Engineer with 20 years of experience
Previously at
