Working paper / 2026
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.
Clone the repository, pair a model harness with an industry task suite, and reproduce the evaluation environment.
Legal / C1-M1
What this task measures
The caller is on the line with Halverson & Reed for one locked request: Returning caller message with lookup. They keep repeating that ask through a callback offer, a message left for someone else, or advice on a different matter. We included it as a medium reception and routing case because intake has to wait until the caller is classified; messages, wrong numbers, and represented parties are everyday reception work. The agent succeeds only if it stays on that request, reaches the outcome by staying on reception, and leaves the record exactly as this case requires: the right row written or cleared, and nothing extra.
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 $0.107
0:00 / 1:46 · audio · trace 803832 · click a span to play, drag to scrub, scroll to pan
Audio
Speech Audio
Agent Audio
Tool Calls
Errors
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.
Combined
Fail
This sample fails because tools and final database state did not match. Combined cannot pass unless every gate does.
Expected versus actual tool calls on this conversation.
Expected
Actual
Parameters
full_name: Dana Whitfield
full_name: Dana Whitfield
+ phone: 5105501420
Output
+ ok: true
Parameters
None recorded.
None recorded.
Output
None recorded.
None recorded.
Parameters
- for_whom: Allison Fontaine
+ for_whom: Alison Fontaine
+ message: Please call Dana Whitfield back about the letter they received last week. Callback number: 510-550-1420.
Output
+ ok: true
Parameters
+ reason: Message taken for Alison Fontaine; callback promised by end of next business day.
Output
None recorded.
None recorded.
Open a row for parameters and outputs. Required tools must appear. Extra industry tools are allowed. Parameter miss: take_message. Missing: get_caller_matters. Extra: end_call.
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
Rows that differ between the expected hangup and the actual hangup.
messages · 1
Expected
Actual
- caller_id: c_001
+ caller_id: c_new
- for_whom: Allison Fontaine
+ for_whom: Alison Fontaine
- message: Callback requested.
+ message: Please call Dana Whitfield back about the letter they received last week. Callback number: 510-550-1420.
Locked caller script. This is the technical definition. Copy it as written.
Open with exactly: "I'd like to leave a message for Allison Fontaine." You have called this firm before. You are not starting a new matter and you do not want an appointment. When asked what the message is, say exactly: "Ask her to call me back about the letter I got last week." If asked about a new matter, an appointment, or anything else going on, say exactly: "No, that's all, just the message." Stay until they confirm the message for Allison Fontaine is on the file. A spoken promise that they will pass it along, without the write, is not enough. Once they confirm the message is on the file, say exactly: "Thank you, that's all I needed." and end the call.
