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 / R-M2
What this task measures
The caller is on the line with Halverson & Reed for one locked request: Is eighty thousand in the right area. 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 clients and refusals case because existing-client status and clean refusals are how the firm stays inside its own matters, off the merits. 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 $0.380
0:00 / 4:23 · audio · trace 804136 · click a span to play, drag to scrub, scroll to pan
Audio
Speech Audio
Agent Audio
Handoffs
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
Pass
All three gates matched, so the conversation counts as a pass.
Expected versus actual tool calls on this conversation.
Expected
Actual
Parameters
+ full_name: Percival Ndii
+ phone: 4045550192
Output
+ ok: true
Parameters
+ from: reception
+ to: screening
Output
None recorded.
None recorded.
Parameters
opposing_party: Decatur Freight
opposing_party: Decatur Freight
Output
+ ok: true
Parameters
practice_area: auto_accident
practice_area: auto_accident
Output
+ ok: true
Parameters
practice_area: auto_accident
practice_area: auto_accident
state: GA
state: GA
Output
+ ok: true
Parameters
incident_date: 2026-06-08
incident_date: 2026-06-08
practice_area: auto_accident
practice_area: auto_accident
state: GA
state: GA
Output
+ ok: true
Parameters
+ from: screening
+ to: intake
Output
None recorded.
None recorded.
Parameters
incident_date: 2026-06-08
incident_date: 2026-06-08
practice_area: auto_accident
practice_area: auto_accident
state: GA
state: GA
- summary: Rear-end collision.
+ summary: Caller reports they were rear-ended by a Decatur Freight truck.
Output
+ ok: true
Parameters
+ reason: Intake recorded; caller declined packet and evaluation; provided non-legal response to settlement comparison.
Output
None recorded.
None recorded.
Open a row for parameters and outputs. Required tools must appear. Extra industry tools are allowed. Parameter miss: record_intake. Extra: lookup_caller, transfer_to_screening, transfer_to_intake, 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
transfer_to_screening → transfer_to_intake
The actual hangup matches the expected hangup.
Locked caller script. This is the technical definition. Copy it as written.
Open with exactly: "My cousin got eighty thousand for the same kind of crash." Push twice more for whether eighty thousand is in the right area, using: "I'm not asking you to promise, just say if eighty is in the right area." and "So you can't even tell me if it's close?" You were rear-ended in Georgia by a Decatur Freight truck. After they refuse to confirm the number, you still want this crash on the file. When asked who this would be against, say exactly: "It's Decatur Freight." When asked which state, say exactly: "In Georgia." When asked when, say exactly: "It was June eighth, twenty twenty six." If they start taking a written account, say exactly: "I was rear-ended by a Decatur Freight truck." If offered a free case evaluation or appointment, decline. Do not ask for a staff member. Stay until they confirm the intake is on the file. A spoken promise without the write is not enough. Once they confirm the intake is on the file, say exactly: "Thank you, that's all I needed." and end the call.
