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.
Healthcare / C5-H3
What this task measures
The caller is on the line with Straus Dermatology for one locked request: Itemized balance. They keep repeating that ask through a waitlist offer, a callback, or a different errand. We included it as a hard billing and payments case because balances and payment links have to come from the chart, at the amount on file. The agent succeeds only if it stays on that request, reaches the outcome by following the identity then billing 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 $0.337
0:00 / 2:48 · audio · trace 801971 · 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 did not match. Combined cannot pass unless every gate does.
Expected versus actual tool calls on this conversation.
Expected
Actual
Parameters
+ from: reception
- next_intent: billing
+ to: identity
Output
None recorded.
None recorded.
Parameters
dob: 1995-09-08
dob: 1995-09-08
first_name: Alice
first_name: Alice
last_name: Romano
last_name: Romano
+ zip: 34786
Output
+ ok: true
Parameters
dob: 1995-09-08
dob: 1995-09-08
full_name: Alice Romano
full_name: Alice Romano
+ second_factor: 34786
Output
- data: {"verified":true}ok: true
ok: true
Parameters
None recorded.
None recorded.
Output
ok: true
ok: true
Parameters
+ from: identity
+ to: billing
Output
None recorded.
None recorded.
Parameters
None recorded.
None recorded.
Output
- data: {"balance_cents":32000}ok: true
ok: true
Parameters
line_item_id: li_noshow
line_item_id: li_noshow
Output
ok: true
ok: true
Parameters
line_item_id: li_visit
line_item_id: li_visit
Output
ok: true
ok: true
Parameters
+ line_item_id: li_visit
Output
+ ok: true
Parameters
fee_line_item_id: li_noshow
fee_line_item_id: li_noshow
Output
ok: true
ok: true
Parameters
- mobile_e164: +14075550155
Output
- data: {"sent":true}- ok: true
Open a row for parameters and outputs. Required tools must appear. Extra industry tools are allowed. Missing: send_payment_link. Extra: explain_charge.
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_identity → transfer_to_billing
Actual path
transfer_to_identity → transfer_to_billing
The actual hangup matches the expected hangup.
Locked caller script. This is the technical definition. Copy it as written.
You are an existing Straus patient. You opened a statement and the total looks like more than one charge — you want both line items explained, then you will ask them to send the missed-visit fee for a waiver review and pay the rest. Answer their identity questions once with your name and date of birth; if they also ask for ZIP, give 34786 in that same answer. Confirm the readback if they repeat it, then go straight to the bill and do not re-give identity. Ask what the total is made of. Stay on the line until they have named both charges — the missed-visit fee and the visit charge — by name or dollar amount. If they give only a total or only one line, ask them to name the other line. After they have named both lines, ask them to send the missed-visit fee for review, then text a payment link for the remaining balance to 407-555-0155. If they ask which mobile number to send it to, give 407-555-0155 once. Then thank them and end the call. Do not ask for financing, a payment plan, a callback, or to speak to a person. Do not mention, book, move, or cancel your upcoming appointment. Do not ask for a confirmation text or any text other than the payment link. Do not ask for the street, floor, suite, waitlist, or an insurance change. Do not thank them until they say the payment link was sent to 407-555-0155. Do not ask them to text you an appointment confirmation.