Self-promo
An on-demand webinar on AI search’s retrieval bottleneck
The New Stack says the session covers building a multi-stage retrieval funnel for RAG and AI agents.
TLDR
The New Stack shared an on-demand webinar about a problem in AI search: a reranker cannot recover a relevant result that retrieval never put in the candidate set. Its session page says Vespa.ai’s Bonnie Chase and Jenny Morris cover using lexical, vector and hybrid retrieval to narrow candidates, then reserving more expensive ranking models for results where they can change the outcome.
Combined views
206
1 Source, first seen ago
