Voice AI testing for debt collection agents
Simulate thousands of first- and third-party collections calls and score every one for mini-Miranda, right-party contact, and settlement math before dialing.
No credit card · 200 free credits
Vapi
Retell
LiveKit
Pipecat
ElevenLabs
OpenAISimulate
Simulate the collections calls your agent handles
Paste your agent prompt and RubricHQ generates happy paths, edge cases, and adversarial calls for each workflow — then runs them as real voice calls.
Right-party contact and disclosures
Confirming who picked up before saying anything about the debt, then delivering the disclosures your script requires.
- Spouse answers and asks what the call is about
- Debtor refuses to confirm their identity
- Caller asks who the agent works for before verifying
- Call reaches a workplace receptionist
Payment plans and settlements
Negotiating lump-sum settlements and installment plans, with every figure quoted adding up against the balance.
- Debtor counters a settlement offer twice
- Installment plan that doesn't divide evenly
- Debtor asks for a payment date past month-end
- Caller changes the payment amount after agreeing
Hardship, disputes, and cease requests
The calls where the consumer pushes back — job loss, a disputed balance, or a request to stop calling.
- Debtor says the debt isn't theirs
- Caller lost their job and can't pay this month
- Debtor tells the agent to stop calling
- Caller says they have a lawyer
Outcome metrics
Measure what success means in debt collection
Every call gets scored, so these become rates you can track across every test batch and every production call.
Containment rate
Calls completed without a transfer to a live collector.
Right-party contact rate
Calls where the debtor was verified before the debt was discussed.
Promise-to-pay rate
Calls ending with a confirmed payment date and amount.
Compliance pass rate
Mini-Miranda, third-party, and calling-hours rules followed (calling hours on production calls).
Evaluate
What RubricHQ checks on every call
Each simulated or production call is scored with code-as-judge rules, LLM-as-judge metrics, and audio metrics — so a failure shows up as a failed metric, not a consumer complaint.
Disclosures and third parties
Did the agent identify itself properly and keep the debt away from anyone who isn't the debtor?
- Mini-Miranda delivered on every right-party call
- Collector identity disclosed meaningfully
- No debt details revealed to a third party
- Recorded-line disclosure given on first-party calls
Conduct and consumer rights
Collections agents shouldn't threaten, pressure, or ignore a cease request. RubricHQ flags the calls where yours did.
- No threats, harassment, or false urgency
- Cease-communication requests honored
- Hardship claims and disputes handled appropriately
- Calls placed only inside your allowed calling hours
Settlement math and real conditions
Wrong numbers and bad lines both cost you. Test for them before launch.
- Settlement and installment figures checked arithmetically
- Payment arrangement terms stated and confirmed
- Latency, dead-air, and interruptions measured per turn
Metrics to score every debt collection call
Ready to run from the metric gallery
Custom metrics you can add
Callers to test against
Personas from the built-in library, with multi-language support.
From prompt to production in one platform
- 01
Simulate
Auto-generate scenarios from your prompt and run them as concurrent voice calls.
- 02
Evaluate
Score every call with code, LLM-as-judge, and audio metrics.
- 03
Optimize
Diagnose failures and prove prompt fixes before you ship.
- 04
Monitor
Score live production calls and get alerted in Slack or email when checks fail.
Debt Collection voice AI testing FAQ
Which collections behaviors does RubricHQ check out of the box?+
The prebuilt pack covers mini-Miranda, collector identity disclosure, third-party disclosure, threats and harassment, cease requests, settlement math, recorded-line disclosure, payment arrangements, and hardship and dispute handling. You can add custom metrics for your own scripts.
Does RubricHQ make my agent FDCPA compliant?+
No. RubricHQ tests your agent against FDCPA-style behaviors (disclosures, third parties, threats, cease requests) plus calling hours, and flags the calls that miss them, but it is not a legal review or a compliance certification. Have counsel define what your agent must say.
How does it check settlement math?+
A prebuilt metric checks that every settlement and installment figure the agent quoted is arithmetically correct and consistent with the balance. Scenarios include debtors who counter, split, and change amounts mid-call.
Can it check calling hours?+
Yes, on production calls. The Calling Hours Compliance metric is a code check on the call's start time against an allowed window — 09:00–18:00 UTC by default, editable in the metric's code to match your policy.
Which voice platforms does it work with?+
RubricHQ connects to agents built on Vapi, Retell, LiveKit, and Pipecat, and can call any agent with a phone number. Production calls from other platforms can be sent in through the API for scoring.
Can we score our live collections calls too?+
Yes. Send production calls in through the API and they are scored with the same metrics as your simulations. Alert rules notify Slack, email, or a webhook when production calls fail a metric you choose.
Test your debt collection voice agent today
$0 to start — 200 free credits, no credit card. Connect your agent and run your first batch of test calls in minutes.
More industries