3x Faster Search: Parallel Test-Time Scaling with Instructed-Retriever-1
Databricks updated Agent Bricks Knowledge Assistant for 3x faster, higher-quality search via parallel test-time scaling.
“Today we're announcing a major update that makes Agent Bricks Knowledge Assistant both faster and higher quality.”
Databricks announced an update to its Agent Bricks Knowledge Assistant that uses parallel test-time scaling with an Instructed-Retriever model to deliver roughly 3x faster search while improving answer quality. It matters as an incremental advance in enterprise RAG/agent retrieval efficiency, but it is a vendor product update rather than a broad industry-shaping signal.