PERIT-S2S · Field support & utilities

Technicians on rooftops and in basements, talking through a fix with a machine running.

Audio comes from technician hotlines, meter reads, safety checks. Scored on task success, with a working field support & utilities operator as the control.

690
Clips in the target held-out set
54h
Audio, never published
5
Languages in the set
1
Working field support & utilities operator as control

Leaderboard

Illustrative figures. Public runs begin when the held-out sets close.
ModelTask success (%)Median turn latencyBarge-in recovery
Human control · sector agentHuman control
93.1 ± 2.0
696ms96.3%
Frontier realtime · lab AFrontier
64.8 ± 2.2
703ms72.6%
Frontier realtime · lab BFrontier
55.2 ± 0.9
919ms58.1%
Frontier multimodal · lab CFrontier
51.1 ± 1.2
1094ms52.2%
Cascaded stack (ASR → LLM → TTS)Specialist
50.0 ± 1.4
1562ms49.4%
Open weights · 8B speechOpen weights
32.1 ± 1.6
1104ms29.4%
Task families

What the set actually asks for.

  • Turn-taking & barge-in
  • Task completion
  • Instruction adherence
  • Disambiguation & repair
  • Escalation & refusal
  • Tone under pressure
Method & limits

How a run on this set is produced.

  1. 01Callers are scripted personas played by trained speakers, not synthesised — including the ones who talk over the agent.
  2. 02Every call is scored against the rubric by two reviewers; a third adjudicates splits.
  3. 038 calls per persona per model. Reported intervals are bootstrapped at 95%.
  4. 04Human control is a licensed agent from that sector handling the same personas blind.

Limits, stated up front: 5 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.