Persistent retrieval cuts costs and speeds up conversational agents

Persistent Retrieval for Conversational Agents: A Study of Retrieval Efficiency

Artificial IntelligenceComputation and LanguageInformation Retrieval

Summary

Conversational agents answer questions by retrieving information in different ways. The authors compared three designs: one that handles retrieval directly, one that uses a helper without keeping memory, and a new helper that remembers past searches and shares evidence with the main agent. Their new design uses fewer computer resources, lowers costs, and gives faster answers over multiple questions. It works by the helper keeping track of what it found before and providing direct evidence instead of summarizing everything.

What this means in practice

Authors

Karthik Ravula, Sathwik Kowtha, Rishabh Verma, Kala Mehak Jain, Gunjit Mittal

Abstract

Conversational agents can retrieve information directly or delegate retrieval to a subagent. This design choice affects how retrieval history is maintained, what context each model call receives, and what information must be re-retrieved in follow-up questions. We compare three designs: skill-based retrieval in the main agent, a non-persistent retrieval subagent that generates a findings response, and our persistent retrieval subagent, which maintains retrieval history across turns and returns retrieved evidence directly to the main agent. Our approach also allows the main agent to request corrections from the same subagent. We evaluate the three designs on enterprise questions posed independently, within shared conversations, and as dependent follow-ups. Across five multi-turn settings, our approach uses 43.3% fewer recorded agent-model tokens and incurs 45.0% lower estimated API cost than skill-based retrieval, while reducing average response time by 9.1%. Compared with non-persistent retrieval, it reduces response time by 15.6 seconds per question, while using 65.1% more tokens and incurring 29.1% higher estimated API cost. Our approach shifts the retrieval subagent from producing synthesized responses to maintaining retrieval state and supplying evidence that the main agent can directly evaluate and use for answer generation.