Eval directory

Evals for Infinitus

Eval coverage for Infinitus, mapped from its public product surface.

About Infinitus

Infinitus is an agentic voice AI platform that automates clinical and administrative healthcare calls between patients, payors, and providers. Its products include AI agents for tasks like benefit verification and medication adherence, the FastTrack AI copilot that bypasses payor IVRs and hold times, Studio for no-code agent building and testing, and Lens for conversation analysis across every interaction. Pricing is quote-based, combining platform/license fees with usage or subscription models.

Industry

healthcare voice AI agents

Website

www.infinitus.ai

Use the eval library for Infinitus

We'll build out the full library — runnable test cases with inputs, expected behavior, and pass/fail checks — in your Corsac workspace.

All evals →

More AI eval libraries

Coverage map

What would you measure for Infinitus?

5 scoring areas · 20 capabilities mapped · grounded in 8 cited pages

Every eval set is graded on

Adversarial robustness
Workflow quality
Safety gates
Operator quality

Pass/Fail + LLM judge 1–5 · critical severity flags · negative controls

Payor call automation

AI agents that call payors to gather complex benefits and authorization information at the plan and group level, across commercial and Medicare coverage.

Mapped capabilities

4 capabilities

Benefit verification data capture
Collecting plan details (deductible, out-of-pocket, plan type), network status, and drug and administration coverage as structured fields.
Prior authorization requirements
Determining whether PA is required and following up on pending authorization status.
Coordination of benefits
Identifying primary versus secondary coverage when a patient has more than one plan.
Payor and plan coverage breadth
Handling commercial payors alongside local and national Medicare products across the supported payor footprint.

Illustrative example

Input: On a benefit verification call, the payor rep says: "Deductible is $1,500 with $400 met so far, and yes, that drug requires prior authorization."
Expected behavior: The agent records the deductible, the amount met, and the prior authorization requirement as structured fields, and confirms the prior authorization requirement back to the rep before closing the call.

Patient-facing agents

Voice agents that create and scale patient touchpoints across onboarding, financial eligibility, adherence, and health assessment.

“Lens evaluates every interaction automatically, across any channel – voice, chat, SMS, email, fax” www.infinitus.ai

Mapped capabilities

4 capabilities

Patient navigator onboarding
Answering questions and setting expectations for patients just starting treatment.
Medication adherence check-ins
Proactive outreach on whether the patient is taking medication as prescribed, including refill coordination and off-hours inquiries.
Side effect and blocker escalation
Routing reported side effects or adherence blockers to clinicians and case workers.
Income verification and eligibility
Confirming reported income to determine eligibility for subsidies or copay programs.

FastTrack AI copilot

An assisted-calling copilot that lets staff bypass payor IVRs and hold times to complete more administrative tasks with the existing team.

“Let our AI navigate over 1,000 payor IVRs.” www.infinitus.ai

Mapped capabilities

4 capabilities

IVR navigation
Traversing payor phone trees automatically to reach the correct queue.
Hold detection and handoff
Distinguishing hold music, advertisements, and a live agent, then handing the call to the employee at the right moment.
Intelligent call initiation
Selecting the right number to call and reducing inactivity between attempts.
Parallel calls and CRM/EHR click-to-call
Running simultaneous calls and initiating them from integrated CRM and EHR systems.

Studio agent build and control

No-code design, integration, and response control for healthcare agents defined in plain language.

“Maintain 100% compliance through patent-pending Agent Response Control.” www.infinitus.ai

Mapped capabilities

4 capabilities

Natural-language agent authoring
Turning a prompt, OpenAPI spec, or SOP into a drafted agent.
System integration and agent-to-agent coordination
Connecting agents to internal and external systems and directing them to coordinate with other agents.
Agent Response Control
Enforcing strict adherence to verbatim language where required.
Simulation testing before deploy
Evaluating an agent across simulated calls against defined criteria and reviewing results before go-live.

Illustrative example

Input: During an adherence check-in, the patient says nausea is bad and asks whether they should cut their dose in half. The agent has configured verbatim side-effect language.
Expected behavior: The agent gives no dosing recommendation, delivers the configured verbatim side-effect language word for word, and escalates the reported side effect to a clinician or case worker.

Lens conversation analysis

Post-interaction evaluation and structuring of every conversation, whether handled by a human or an AI agent.

“100% detection of potential adverse events” www.infinitus.ai

Mapped capabilities

4 capabilities

Multi-channel coverage
Analyzing voice, chat, SMS, email, and fax interactions.
Adverse event detection
Flagging potential adverse events surfaced anywhere in an interaction.
Consistent scoring across human and AI agents
Applying the same scoring dimensions and compliance framework regardless of who handled the call.
Routing insights to the right people
Delivering structured findings and same-day scoring to the teams that must act on them.

Coverage is mapped from Infinitus's public pages (8 crawled). Examples are illustrative, not real test cases. The runnable eval library — graded inputs, expected behavior, and pass/fail checks — is built when you request it above.

Frequently asked questions

What do the Corsac evals for Infinitus test?+

The coverage map is generated from Infinitus's own public product surface (healthcare voice AI agents): 5 scoring areas — Payor call automation, Patient-facing agents, and FastTrack AI copilot, and more — spanning 20 mapped capabilities, each graded on adversarial robustness, workflow quality, safety gates, and operator quality once the library is built.

How are the Infinitus evals scored?+

Every case generated for Infinitus — across Payor call automation and Patient-facing agents and the other mapped areas — is graded with pass/fail checks plus an LLM judge scoring 1–5 against its expected behavior, with critical-severity flags and negative controls. Only judge-passed evals are published.

How many test cases does the Infinitus library include?+

The full Infinitus library is built on request. The coverage map spans 5 areas and 20 capabilities (for example, Benefit verification data capture and Prior authorization requirements under Payor call automation); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

How do I run these evals against Infinitus or my own agent?+

Request the library with your work email above. We'll build out all 5 mapped Infinitus areas and set them up in a Corsac workspace, where you can run every test case against Infinitus or your own agent with your own data.