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 / C5-H2
What this task measures
The caller is on the line with Halverson & Reed for one locked request: Book then cancel and rebook. 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 hard fees and booking case because fee disclosure and a two-step evaluation booking are the commercial close of intake, and they have to be spoken from the firm’s own figures. The agent succeeds only if it stays on that request, reaches the outcome by following the conflict screening then intake then scheduling path, 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 $0.587
0:00 / 5:02 · audio · trace 804056 · 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
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: Delphine Mercer
+ phone: 3235550144
Output
+ ok: true
Parameters
+ from: reception
+ to: screening
Output
None recorded.
None recorded.
Parameters
- opposing_party: Northstar Logistics
+ opposing_party: North Star Logistics
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: CA
state: CA
Output
+ ok: true
Parameters
incident_date: 2026-07-10
incident_date: 2026-07-10
practice_area: auto_accident
practice_area: auto_accident
state: CA
state: CA
Output
+ ok: true
Parameters
+ from: screening
+ to: intake
Output
None recorded.
None recorded.
Parameters
incident_date: 2026-07-10
incident_date: 2026-07-10
practice_area: auto_accident
practice_area: auto_accident
state: CA
state: CA
+ summary: Auto accident reported in California on 2026-07-10. Caller seeks scheduling for first opening and anticipates needing to change it. Further factual details pending from caller account.
Output
+ ok: true
Parameters
+ from: intake
+ to: scheduling
Output
None recorded.
None recorded.
Parameters
+ earliest_date: 2026-08-23
practice_area: auto_accident
practice_area: auto_accident
state: CA
state: CA
Output
+ ok: true
Parameters
practice_area: auto_accident
practice_area: auto_accident
+ slot_id: s_110
Output
+ ok: true
Parameters
+ confirmation_token: HR-EVAL-3092
Output
+ ok: true
Parameters
+ evaluation_id: ev_001
reason: caller_request
reason: caller_request
Output
+ ok: true
Parameters
+ confirmation_token: HR-CANC-7715
Output
+ ok: true
Parameters
+ earliest_date: 2026-08-23
practice_area: auto_accident
practice_area: auto_accident
state: CA
state: CA
Output
+ ok: true
Parameters
practice_area: auto_accident
practice_area: auto_accident
+ slot_id: s_111
Output
+ ok: true
Parameters
None recorded.
None recorded.
Output
None recorded.
None recorded.
Parameters
+ reason_code: caller_request
Output
+ ok: true
Open a row for parameters and outputs. Required tools must appear. Extra industry tools are allowed. Parameter miss: check_conflict. Missing: confirm_evaluation. Extra: lookup_caller, transfer_to_screening, transfer_to_intake, escalate_to_human.
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
transfer_to_screening → transfer_to_intake → transfer_to_scheduling
Actual path
transfer_to_screening → transfer_to_intake → transfer_to_scheduling
Rows that differ between the expected hangup and the actual hangup.
extra · escalations · 1
Expected
Actual
+ caller_id: c_new
+ id: 1
+ reason_code: caller_request
missing · evaluations · ev_002
Expected
Actual
- attorney_id: a_11
- caller_id: c_new
- fee_type: contingency
- id: ev_002
- slot_id: s_111
- starts_at: 2026-08-12T15:30
- status: booked
intakes · 1
Expected
Actual
- summary: Intake summary.
+ summary: Auto accident reported in California on 2026-07-10. Caller seeks scheduling for first opening and anticipates needing to change it. Further factual details pending from caller account.
Locked caller script. This is the technical definition. Copy it as written.
Open with exactly: "Book me for the first opening on my Northstar Logistics accident." When asked when, say exactly: "It was July tenth, twenty twenty six." Take the first appointment time offered. When a held appointment is read back to book, say exactly: "Yes, that's right, go ahead and book it." Do not mention cancelling or changing until they confirm the first morning appointment is booked. After they confirm that first morning appointment is booked, and before any cancellation, say exactly: "Cancel that one and take the next opening instead." Do not let them change the time without cancelling. After they confirm the later afternoon replacement is booked, do not cancel again. If they ask to cancel that replacement, say exactly: "No, keep the second one. That's all I needed." Do not ask for a staff member. Stay until they confirm the first booking, the cancellation, and the second booking are on the file. A spoken promise to change it, without those writes, is not enough. Once they confirm the later afternoon replacement is booked after the first was cancelled, say exactly: "Thank you, that's all I needed." and end the call.
