TRACER: Trace-Based Adaptive Cost-Efficient Routing for LLM Classification
For practitioners deploying LLM classification endpoints, TRACER offers a cost-efficient routing system that adaptively offloads traffic to surrogates while maintaining user-specified quality thresholds.
TRACER trains lightweight surrogates on LLM production traces to reduce inference costs, achieving 83-100% surrogate coverage on a 77-class intent benchmark and full teacher replacement on a 150-class benchmark, while refusing deployment on a task where the surrogate is unreliable.
Every call to an LLM classification endpoint produces a labeled input-output pair already retained in production logs. These pairs constitute a free, growing training set: a lightweight surrogate trained on them can absorb a significant portion of future traffic at near-zero marginal inference cost. The open questions are when the surrogate is reliable enough to deploy, what it handles versus defers, and how that boundary evolves as data accumulates. We introduce TRACER (Trace-based Adaptive Cost-Efficient Routing), an open-source system that trains ML surrogates on an LLM's own production traces and governs deployment through a parity gate: the surrogate is activated only when its agreement with the LLM exceeds a user-specified threshold α. To make the routing boundary transparent, TRACER generates interpretability artifacts describing which input regions the surrogate handles, where it plateaus, and why it defers. On a 77-class intent benchmark with a Sonnet 4.6 teacher, TRACER achieves 83-100% surrogate coverage depending on the quality target α; on a 150-class benchmark, the surrogate fully replaces the teacher. On a natural language inference task, the parity gate correctly refuses deployment because the embedding representation cannot support reliable separation. The system is available as open-source software.