LLM eval & observability platform — eval / observability platform

LLM eval & observability platform — eval / observability platform

Opportunity: Heaviest demand-query cluster in the dataset (>16 issues across Langfuse, Phoenix, Ragas, DeepEval). Users are now operating these at production scale — bugs are about ClickHouse migrations, S3 replay, OTEL ingestion, multi-agent cost aggregation. That is the classic 'crossed-the-chasm' bug profile.

Demand-side hot, supply-side fragmented. The bug profile (token-cost accounting, OTEL semantics, eval reproducibility) hints at the consolidation wedge: whichever vendor first nails multi-agent cost + eval correctness wins the category.