Many of our customers run workloads with speaker attribution, with the strictest demands on both quality and speed.
We partnered with the team at
@pyannoteAI to deliver both: some of the highest-quality diarization models on the market, with 3.2x higher throughput and 9.6x lower latency.