[Feedback] Experimental release of Agentic Search in 3.2

We released Agentic Search as an experimental feature in OpenSearch 3.2 release. It is a new query type proposed in OpenSearch that triggers an agent-driven workflow for query understanding, planning and execution. Instead of hand-crafting DSL, you supply a natural language question and an agent id. The agent then executes a query planning tool to produce OpenSearch DSL, run it via agentic query clause and return the search results.

As mentioned, this feature is experimental as of now. We’d love to gather some feedback from you on this feature, to help us understand your need around it, and plan our next steps accordingly. Any comments/suggestions will be highly appreciated!

2 Likes

Hi,

I am trying to evaluate agentic search in opensearch for NLQ to Opensearch query , could you please let me know if there is any golden NLQ and SQL pairs /dataset available which I can use to evaluate the accuracy/correctness of DSL query generated ? How was agentic search evaluated ? any evaluation test cases that are available for verification ?