Introducing Exa ConnectLearn more

WEBINAR

← Back to all webinars

Improving RL Outcomes with Exa

Wednesday, September 16, 2026 · 10:00–11:00 AM PT

Web data shapes every RL training run, yet the retriever is one of the hardest parts for search-agent teams to tune.

Register for the webinar

In this session, Josh Yutsever from Exa's ML team shares our research on why agents trained with Exa learn faster, use fewer tokens and search calls, and score higher than agents trained on traditional search methods.

What you'll learn

  • How to set up a controlled A/B of retrieval backends inside an RL loop
  • Where token savings come from with web search data from Exa
  • How to read pass@k across k to tell real gains from probability-mass concentration

Who should attend

  • ML engineers and researchers post-training agents with RL
  • Teams building deep research or search agents on live web data
  • Anyone choosing a retrieval backend for a training pipeline and wondering whether it matters