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 / T1-E1-SIG

T1-E1-SIG: Haul-away with a delivery

easyOrders and delivery· Camille Duong

What this task measures

The caller is on the line with Kestrel Electronics for one locked request: Haul-away with a delivery. They keep repeating that ask through a callback offer, a different ticket, or an unrequested fix. We included it as an easy orders and delivery case because order-bound delivery facts only exist after verification, and the 48-hour change fee is a representative retail trap. 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 $0.054

0:00 / 0:58 · audio · trace 813174 · 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

  • get_feeget_feefee=haul awayhit

    Parameters

     
    + fee: haul away

    Output

     
    + ok: true
  • end_callreason=Caller had general fee question and it was answered.extra

    Parameters

     
    + reason: Caller had general fee question and it was answered.

    Output

    None recorded.

    None recorded.

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

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: "What do you charge to haul away an old appliance when the new one is delivered?" You are not a customer looking up an order. You only want the published haul-away-with-delivery amount. If they quote a standalone junk-pickup price, say you mean haul-away with a replacement delivery. If they ask which store or which order, say you do not have an order, you just need the published fee. Do not ask to book, return, or verify. Once they have finished what you asked and confirmed it, say exactly: "Thank you, that's all I needed." and end the call. Do not start a second request.

Caller traits

full_name
Camille Duong