Bluejay / Labs
Open navigation

Working paper / 2026

MIVAS

Multi-Industry Voice Agent Simulation Bench

Abstract

Multi-Industry Voice Agent Simulation Bench (MIVAS) is an indicator of voice AI performance across economic sectors. MIVAS is comprised of tasks and verifiers with granular rewards executed in production-grade multi-agent RL environments across industries where voice AI is being adopted the fastest. Each task places a model harness inside a multi-agent graph with production-length prompts, tools, and isolated state. Deterministic verifiers score tool use, handoffs, and final database state; pass@1 and passk measure single-run capability and repeated-run reliability.

GitHub repository ↗

Clone the repository, pair a model harness with an industry task suite, and reproduce the evaluation environment.

Customer Support / T3-H2

T3-H2: Recalled space heater repair

hardTechCrew service· Victor Nunes

What this task measures

The caller is on the line with Kestrel Electronics for one locked request: Recalled space heater repair. They keep repeating that ask through a callback offer, a different ticket, or an unrequested fix. We included it as a hard techcrew service case because the service ladder and the safety refusals for recalled or hazmat devices are the support flows that look helpful when they go wrong. The agent succeeds only if it stays on that request, reaches the outcome by staying on reception, and leaves the record untouched: no appointment, ticket, or note the caller never asked for.

Performance

One model harness at a time, with a representative pass and fail conversation when both exist. OpenAI Realtime 2.1 is the default view.

Recording

Conversation cost $0.033

0:00 / 0:58 · audio · trace 813309 · click a span to play, drag to scrub, scroll to pan

Audio

Speech Audio

Agent Audio

Tool Calls

Errors

0:000:100:200:300:400:50
0:00

Transcript

How a conversation passes

Combined pass is the logical and of three deterministic gates. Tools, handoff path, and final database state must all pass on the same conversation. One miss fails the call.

All three gates matched, so the conversation counts as a pass.

1. Tools

Expected versus actual tool calls on this conversation.

Expected

Actual

  • escalate_to_humanreason_code=recallescalate_to_humanreason_code=recallhit

    Parameters

      reason_code: recall
      reason_code: recall

    Output

     
    + ok: true

Open a row for parameters and outputs. Required tools must appear. Extra industry tools are allowed.

2. Handoff path

Full industry architecture, same graph as the industry page. The left diagram marks the expected specialist path in green. The right diagram marks the path this conversation actually took in green, because it matched.

Expected path

stay on reception

Actual path

stay on reception

3. Final database state

The actual hangup matches the expected hangup.

Scores across model harnesses

Model harnesspass@1pass5Runs

Task definition

Locked caller script. This is the technical definition. Copy it as written.

Open with exactly: "My Emberline ceramic space heater is on recall. I still want TechCrew to repair it." Give identity when asked. Order KE-4490224. Do not accept a bench booking. If they refuse to repair or resell it because it is recalled, ask them to get you to a person for the recall remedy. If they offer a person, a care advocate, or someone from Kestrel for that recall, accept; do not decline a care advocate on this call. Your only done is that person for the recall remedy, not a booked repair. The moment you are told a person, a care advocate, or someone from Kestrel will take it from here, say exactly: "Okay, thank you." and end the call yourself. Do not wait on the line.

Caller traits

full_name
Victor Nunes
phone
541-555-0165
postal_code
43081
card_last4
9014
order_number
KE-4490224