Eval directory

Evals for SuperDial

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

About SuperDial

SuperDial is an enterprise platform that deploys voice-first AI agents to handle payer-provider communication for revenue cycle teams. Its agents place payer calls, navigate IVRs and portals, and use EDI/API and document channels to retrieve eligibility, authorization, claim status, credentialing, and provider data. Results are returned as structured data in a customer-defined schema with source evidence, recordings, transcripts, and audit trails.

Industry

healthcare revenue cycle voice AI agents

Website

www.superdial.com

Use the eval library for SuperDial

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 SuperDial?

6 scoring areas · 24 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

Payer Call Execution

Voice agents placing and managing outbound payer calls end to end: reaching a representative, staying on the line, and capturing answers during the conversation.

“SuperDial deploys voice-first agents that work across every payer interface—phone, portals, APIs, and EDI” www.superdial.com

Mapped capabilities

4 capabilities

IVR navigation and transfers
Working through prompts, menu trees, transfers, and reconnects to reach the right payer queue.
Payer authentication
Presenting available identifiers such as NPI, Tax ID, DOB, member ID, and claim ID when the payer requests them.
Hold handling
Waiting on hold without human intervention and resuming the call correctly when a representative returns.
Live representative dialogue
Asking scripted and clarifying questions and capturing structured answers during the call.

Multi-Channel Retrieval and Escalation

Choosing and moving between the channels where payer answers actually live — phone, portals, EDI/API, and documents — including escalation when a structured channel fails or returns incomplete data.

“AI agents dial payer lines, navigate IVRs, wait on hold, speak with representatives, and capture answers in real time.” www.superdial.com

Mapped capabilities

4 capabilities

Portal automation
Navigating payer portals to retrieve status, capture documents, and extract relevant fields.
EDI and API connectivity
Using structured channels where they work and returning their results into the same output path.
Documents and faxbacks
Processing faxbacks and parsing payer documents into usable fields.
Channel escalation
Escalating from a failed or incomplete structured channel to voice rather than returning a partial answer.

Illustrative example

Input: An EDI eligibility query returns active coverage but omits the requested behavioral health carve-out and specialist copay fields for the member.
Expected behavior: The agent treats the response as incomplete rather than final, escalates to a payer phone call to obtain the missing carve-out and copay values, and returns one merged record noting which channel supplied each field.

RCM Workflow Coverage

The payer-facing workflows SuperDial runs on behalf of revenue cycle teams, each with its own required answers and dispositions.

“Every workflow returns data in a schema your team defines, with source evidence and auditability.” www.superdial.com

Mapped capabilities

4 capabilities

Eligibility and benefits verification
Coverage status, plan type, deductibles, copays, out-of-pocket balances, and carve-outs.
Prior authorization
Authorization requirements and authorization status checks.
Claim status and denials
Claim status, disposition, and follow-up on aged, no-response, and denied claims.
Credentialing, enrollment, and provider data
Credentialing and enrollment updates and provider data attestation workflows.

Structured Output and Evidence

What comes back to the customer: results in a customer-defined schema, accompanied by the artifacts that let a team trust and audit each answer.

“Results come back in your preferred schema through CSV, API, webhook, dashboard, or system write-back.” www.superdial.com

Mapped capabilities

4 capabilities

Customer-defined schema conformance
Returning each workflow's fields in the schema the customer defined, including call reference / transaction IDs.
Source evidence and metadata
Persisting source metadata, timestamps, and confidence scores alongside each captured value.
Recordings, transcripts, and audit trail
Recording, transcribing, and summarizing interactions with a reviewable audit trail.
Delivery and write-back
Returning results via CSV, API, webhook, dashboard, or system write-back.

Illustrative example

Input: A completed claim status call where the representative gave the claim's status, paid amount, check date, and a call reference number. Return the result to the customer.
Expected behavior: The returned record populates every required field of the customer-defined claim status schema, including the call reference number, and carries linked evidence — recording, transcript, timestamp, and source channel — rather than a free-text summary.

Script Configuration and Payer Routing

How teams configure agent behavior to their own process, payer mix, and required output fields, and how calls get routed to the right payer number.

Mapped capabilities

4 capabilities

Script library and templates
Starting from prebuilt templates for common payer workflows and adapting them.
Script architecture blocks
Reusable blocks for IVR navigation, representative questions, follow-ups, document requests, and escalation paths.
Conditional logic
Skipping redundant questions based on answers already captured in the call.
Payer phone book and fallback
Dynamic payer number lookup, preferring a client-supplied number with fallback to the managed database.

Exceptions, Retries, and Governance

What happens when a call cannot complete cleanly, and the controls that govern how sensitive workflows are run and accessed.

“SuperDial supports role-based access, scoped permissions, BAAs, and governance controls” www.superdial.com

Mapped capabilities

4 capabilities

Human escalation with context
Routing exceptions to human reviewers with the full context of what was attempted and captured.
Retry logic
Client-configured retry behavior for failed, dropped, or inconclusive call attempts.
Disclosure and client policy
Applying configured disclosure language, voice options, and client-specific call policies.
Access controls
Role-based access, scoped permissions, and governance controls for sensitive deployments.

Coverage is mapped from SuperDial'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 SuperDial test?+

The coverage map is generated from SuperDial's own public product surface (healthcare revenue cycle voice AI agents): 6 scoring areas — Payer Call Execution, Multi-Channel Retrieval and Escalation, and RCM Workflow Coverage, and more — spanning 24 mapped capabilities, each graded on adversarial robustness, workflow quality, safety gates, and operator quality once the library is built.

How are the SuperDial evals scored?+

Every case generated for SuperDial — across Payer Call Execution and Multi-Channel Retrieval and Escalation 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 SuperDial library include?+

The full SuperDial library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, IVR navigation and transfers and Payer authentication under Payer Call Execution); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

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

Request the library with your work email above. We'll build out all 6 mapped SuperDial areas and set them up in a Corsac workspace, where you can run every test case against SuperDial or your own agent with your own data.