Fine-tune a search agent with multi-turn RL on Amazon SageMaker AI

What happened

No base model arrives knowing your tools or your environment. Fine-tuning teaches a small search agent your tools and environment, giving it the reliability of a frontier model at lower latency (the delay between asking for something and getting it) and cost.

Search agents powered by large language model (the kind of AI system trained on text to produce text)s (LLMs) are transforming how enterprises retrieve information. Rather than requiring users to craft the perfect query, a search agent autonomously decides what to search for, which retrieval strategy to use, and when to stop searching.

It does this across multiple rounds of interaction, refining its approach based on what it has already retrieved. However, getting this multi-step behavior to work well is hard.