Haystack US 2026 will be held during the week of 4th May in Charlottesville, Virginia.

At Haystack, we want to hear from practitioners from around the world who are working on hard search & AI problems. Search as a product and as an enabling technology is on the cusp of a transformative era. The era of AI-powered search and generative AI. Our attendees want to know everything about this journey, from the big picture business decisions down to the nuts and bolts of selecting an algorithm or model to realize a technical goal. Your audience will generally be search professionals, search product owners, search product managers, AI and relevance engineers, and other developers.

This year, we’ll be looking for submissions across two specific areas of interest.

• AI-powered search from a product perspective – The big picture choices, and the pitfalls along the way.
  • Especially interesting are end-to-end case studies that elaborate on the business decisions that drove technology adoption
  • Also interesting are case studies of the evolution of search as a product in an AI-powered world
• AI-powered search as a technology – vector search, embeddings, LLMs, Retrieval Augmented Generation, Context Augmented Generation, search result quality evaluation, complex system evaluation, including RAG evaluation, and so on.

Talks at Haystack are 40 minutes long. Please include time for Q&A.

All speakers who are accepted will receive a 50% discount on the conference ticket either by discount code or via a refund (we try very hard to keep ticket prices as low as possible, plus we know some speakers pay for their ticket in advance in full anyway, so we feel this is an acceptable compromise).

Submissions will be reviewed by multiple reviewers. Reviewing will be conducted on a double-blind basis, with identifying information removed, by a volunteer group drawn from the search relevance community. The top-rated submissions will then be used to create the Haystack conference program.

Submissions close on 2026-02-22 23:59 (America/New_York), 1 month from now.