Blog
Notes on speech-to-text from the people building it — how the metrics are actually measured, and what we learned getting the numbers we publish.
· 6 min read
Why two providers can both claim the lowest diarization error rate
Diarization error rate depends on a tolerance window called the collar, and a figure quoted without one is not comparable to anything. What the collar does, why 0.25 s and 0 give different answers, and how to read a DER claim.
Read →· 5 min read
Batch or real-time: choosing the wrong one costs accuracy or costs latency
Real-time and batch transcription are different products with different accuracy ceilings, not fast and slow versions of one thing. When streaming is genuinely required, and when it quietly costs you quality for nothing.
Read →