PERIT-STT · Retail & e-commerce

Order status, returns and refunds, with catalogue names that no general model has seen.

Audio comes from order support, returns desks, seller helplines. Scored on word error rate, with a working retail & e-commerce operator as the control.

1,340
Clips in the target held-out set
92h
Audio, never published
8
Languages in the set
1
Working retail & e-commerce operator as control

Leaderboard

Illustrative figures. Public runs begin when the held-out sets close.
ModelWord error rate (%)Entity accuracyDiarization error
Human control · sector transcriptionistHuman control
2.2 ± 1.4
98.8%3.1%
Frontier ASR · lab AFrontier
8.1 ± 2.1
91.3%8.2%
Specialist vendor ASRSpecialist
9.0 ± 1.8
89.3%8.4%
Frontier multimodal · lab CFrontier
9.8 ± 1.5
84%12.4%
Frontier ASR · lab BFrontier
10.5 ± 0.9
87.9%11.9%
Open weights · 8BOpen weights
20.6 ± 2.0
70.4%18%
Task families

What the set actually asks for.

  • Verbatim transcription
  • Entity capture (IDs, amounts, names)
  • Speaker diarization
  • Code-switching
  • Noisy & far-field audio
  • Domain vocabulary
Method & limits

How a run on this set is produced.

  1. 01Audio is held out and never published. A 60-clip development split per sector is open for calibration.
  2. 02Reference transcripts are written by two annotators independently; a senior reviewer adjudicates every disagreement.
  3. 03Each clip is run 3 times per model. Reported intervals are bootstrapped at 95%.
  4. 04Human control is a working transcriptionist from the same sector, timed and paid at market rate.

Limits, stated up front: 8 languages only, single-channel audio, and no cross-sector transfer. The control is one operator per sector, so the human line carries its own error bar — it is a working standard, not a ceiling.