RCM & Prior Auth AI
S

SuperDial

SuperDial sells voice AI agents that work payer requests for revenue cycle teams: benefits verification, prior authorization status, claim status, credentialing and enrollment, and provider data checks. An agent dials the insurer, moves through the phone menu, waits on hold and asks a representative only for the fields still missing, and when a payer offers the answer through an electronic transaction, an API, a portal or a faxback, it takes that path first. Each result comes back as structured fields with the source, a timestamp, confidence markers on key fields, a call summary and, for calls, the recording and transcript. Requests the agent cannot finish return to staff marked unable to complete, with the reason.

Work arrives by CSV, a work queue export, the Requests API with webhooks or a practice system trigger, and results go back by API, export, a portal dashboard, work queue notes or write back into systems such as athenahealth, AdvancedMD and Open Dental. United Medical Monitoring retrieved the required claim information on 83% of more than 5,900 calls over just over three months, and MBW RCM, an outsourcer with more than 700 billing staff, reports 50% lower call costs at more than 1,000 calls a month. In November 2026 SuperDial passed 10 million payer calls and launched Agent-to-Agent RCM, which counts resolved requests rather than calls. Sam Schwager and Harrison Caruthers founded the San Francisco company as SuperBill in 2021, and its $15 million Series A in June 2025 was led by SignalFire.

AI Health Index verifiedNovember 20, 2026
Compare SuperDial with other vendors
Founded
2021
Headquarters
San Francisco, California
Categories
rcm-and-prior-auth, healthcare-admin-automation
Indexed Products
SuperDial Voice Agents, Agent-to-Agent RCM, SuperDial Requests API
Buyer Segments
Revenue cycle management firms, Provider groups, Health systems, Dental service organizations, Payer operations teams, Digital health platforms
Assessment

Capability Axes

An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read

AI Capability
AA on AI CentralityThe artificial intelligence is the product. Remove the model and there is nothing left to sell.
Vendor Published

The agent is the product: it does the work a biller does on a payer line. It dials the insurer, works through the interactive voice menu, waits for a representative, asks for a defined set of answers and writes them back as structured fields, carrying forward what it already knows so it asks only for what is missing.

Electronic transactions, payer APIs, portals and faxbacks answer the same request when a payer offers the data digitally, and Agent-to-Agent RCM extends the exchange to payers' own voice automation. The Requests API, CSV uploads and dashboards exist to feed that agent and read its output, so without the model there is nothing left to buy.

AA on Autonomy and Oversight ModelWhat the system may do and what it may not do are both published, with escalation thresholds, override paths and the conditions that route a case to a person.
Vendor Published

A request ends one of three ways: completed with its fields, returned as an input error such as a wrong member ID, NPI or claim number for staff to fix and resubmit, or marked unable to complete with the reason. When a task needs subjective judgment, a payer escalation or a field the payer will not release, the agent routes a packet to the right person or team showing what it tried, what it captured, what is missing and a recommended next step, and a human fallback team completes calls the agents cannot. Customers configure question blocks, escalation rules, disclosure language and retry logic, see confidence scores on key fields and can switch off the practice system connection.

BB on Model and Technology TransparencyThe approach or the suppliers are named without the version and update discipline behind them. Naming a supplier is the entry to this band both here and on Model Supply Chain Disclosure, which ask different questions of the same disclosure: who receives the data, and what produces the output.
Vendor Published

Each request runs through channel selection, voice conversation and field extraction, and the Requests API defines inputs and outputs in a schema so results from a call, a portal or an electronic transaction come back in one format. Deepgram and AssemblyAI handle speech recognition, Cartesia supplies voice, OpenAI and Anthropic are the AI production platforms, and Cerebrium and Baseten host model inference.

Model names and versions are not given, nor which model handles navigation, conversation or extraction, nor how a model change reaches a customer's workflow. Confidence scores on key fields are the one output level signal a customer sees of how sure the model was.

BB on Model Supply Chain DisclosureSubstantial partial disclosure, or a chain that is structurally short: an in house build, a cleared model that cannot be quietly swapped, or a deployment where the transfer does not occur at all. Naming the model provider exits the band below into this one; the axis rises from here on the completeness of the party list and on the terms that govern data once it arrives.
Vendor Published

The trust center's subprocessor list names 16 parties, all in the United States: Deepgram, AssemblyAI and Cartesia for speech and voice, OpenAI and Anthropic as AI production platforms, Cerebrium and Baseten for inference hosting, Google Cloud Platform for infrastructure, Telnyx and Jambonz for messaging and telephony, ClickHouse for analytics, Cloudcruise for healthcare workflow automation, and Google Workspace, Slack, Linear and Notion for internal work.

Protected health information is processed under each customer's business associate agreement. The list stops at names: whether OpenAI, Anthropic and the speech providers retain audio, transcripts or submitted fields, and whether any may train on them, is not set out.

BB on Clinical and Operational EvidenceNamed deployments with dated outcome figures and enough method to test them, or published research short of independent validation.
Vendor Published

United Medical Monitoring, which bills for neuromonitoring, ran more than 5,900 payer calls over just over three months, and the required claim information came back on 83% of requests. MBW RCM, an outsourcer with more than 700 billers, coders and receivables staff serving over 40 customers, reports 50% lower costs after replacing manual calls, at more than 1,000 calls a month, with no dates given.

Other customer figures are headlines only: more than $12 million in billed claims worked for Apex IONM, more than 100 hours a month saved at Mentaya and 1,000 calls a month at CBS Medical Billing. The November 2026 timing of about four minutes per attempt against payer voice automation, and 17 minutes or more to reach a representative, is SuperDial's own operating data.

CC on AI Safety and PHI StewardshipGeneral assurances of privacy and security that do not answer the questions artificial intelligence raises: what is retained, what reaches a model, and what happens to it there.
Vendor Published

Requests carry what a biller would read to the insurer, including NPI, tax ID, date of birth, member ID and claim number, and results can include audio, a transcript and a structured record. Protected health information moves under each customer's business associate agreement, and all 16 subprocessors sit in the United States, with OpenAI, Anthropic, Deepgram, AssemblyAI and Cartesia named for AI and speech processing.

Retention policies and model training restrictions are options an enterprise customer can configure, and the trust center lists general retention procedures and deletion of customer data when a customer leaves. No default retention period for recordings or prompts, and no standing commitment against training on customer data, is stated.

Regulatory and Compliance
BB on HIPAA and BAA PostureBusiness associate status is stated and supported by a substantive privacy document, with the agreement or its scope not fully published. For a vendor outside the United States, an equivalent regime documented to this depth grades here. A vendor whose product does not process protected health information grades here too when it says so plainly and explains the scope, with a service privacy document behind it: stating a position a buyer can rely on is the posture this axis grades, and the band above is closed to it because there is no agreement to publish.
Vendor Published

Every customer signs a business associate agreement before any protected health information is shared, and the site's badges add HIPAA compliance and BAA availability. SuperBill, Inc., the legal entity, has a product and service privacy notice explaining that covered entities or other business associates direct its processing under that agreement, and that corporate customer data stays governed by the customer's contract, including information used to retrieve a claim or verification response. Individual access requests go to the organization that controls the data. The agreement's text, the tier it applies at and the terms binding the 16 subprocessors to it are not public.

AA on Security Certifications and Trust CenterCertifications named with their type and version and presented as retrievable artifacts, usually through a trust portal a buyer can open without asking.
Vendor Published

The HITRUST e1 certificate opens to view and download on the trust center without a request, while the HITRUST e1 report and the 2024 and 2025 SOC 2 Type II reports need an access request. Listed controls cover restricted access to encryption keys, production systems, databases and networks, multifactor authentication for remote access, encryption at rest and in transit, log management, annual penetration testing with a remediation plan, annual control self assessments, tested disaster recovery plans and cybersecurity insurance.

Customers get single sign on, multifactor authentication, role based access and scoped permissions, with AES 256 bit encryption at rest. The e1 is HITRUST's entry level assessment, narrower in scope than its i1 and r2.

CC on FDA and Regulatory StatusNo device claim is made and the product is scoped accordingly. Most administrative and operational products sit here and are not penalized for it, because this axis grades the appropriateness of the positioning rather than possession of a clearance. A stated limit on use, such as research use only or a clause that the product gives no medical advice, beside marketing that sells the product for clinical decisions also sits here, with both quoted, because the buyer has to ask which one governs.
Vendor Published

The agents retrieve administrative facts a payer already holds: eligibility and benefits, allowed amounts, claim status and denial reasons, prior authorization status, and credentialing details such as panel status, effective dates and missing forms. A prior authorization check reports the payer's decision rather than making one, and nothing the agent returns is a clinical recommendation.

Configurable question blocks set what is asked, and the answers land in revenue cycle fields for billing and enrollment staff. The product makes no device claim and needs none for that scope. SuperDial also states no regulatory position of its own, so the administrative scope rests on what the product plainly does.

BB on AI Governance and Bias DisclosureA governance framework with named process behind it, such as certification to an artificial intelligence management standard, or material written for a customer own review committee to evaluate the product with. A framework covering the vendor whole portfolio counts when its scope reaches this product; what it cannot supply is the product specific evaluation the band above asks for. Published results within the one group where the performance of this product is most at risk also sit here: evaluation on the question this axis asks, short of the comparison across groups the band above requires. A medical device clearance or certification, and the quality system behind it, does not reach this band on its own; it is graded under regulatory status. A regulator reviewed plan governing how the model may change does reach it.
Vendor Published

Pilots run on the customer's own payer mix, usually over 30 to 90 days, with the customer defining the output schema and what counts as a completed request. MBW RCM began with a handful of test calls, and its operations leaders approved the call scripts for each claim and benefits scenario before volume grew.

Every result carries its source, field confidence markers and a status code, and incomplete requests can be checked against the recording and transcript, which gives an operations team material to evaluate the agent against its own cases. No results compare accuracy across payers, call types or representatives' accents, and no outside review of how the agents behave is named.

CC on AI Liability and RecourseMechanisms exist that let someone challenge an output, such as audit trails, source traceability or review before commit, with nothing standing behind the output and no route for the harmed party.
Vendor Published

Recordings, transcripts and request logs let a billing team trace each answer to the call or portal page it came from and see where an extracted field differs from what the representative said. Input errors such as a wrong member ID, NPI or claim number are flagged for correction and resubmission. Under Agent-to-Agent RCM, SuperDial takes responsibility for resolving a request across channels, which describes the job rather than a remedy.

SuperBill's service terms disclaim liability for content errors and cap liability at the greater of fees paid or $100 where the law allows, with a dispute procedure. No warranty, credit or rework obligation attaches to a wrong benefit, claim status or authorization answer.

Integration and Deployment
BB on EHR and Interoperability DepthNamed systems with read access or one directional writing, or standards support with named deployments behind it.
Vendor Published

A scheduled appointment, an aging claim or a pending authorization in the practice system can trigger work, and results return by API and webhooks, CSV or PDF files, the portal, work queue notes or write back into the system. athenahealth customers enable SuperDial from the athenahealth Marketplace, DrChrono users authorize it inside their account, and NextGen works with or without NextGen's paid API.

Tebra receives a sorted results file, CureMD feeds work through its reports, Dentrix Ascend runs with Henry Schein One, and AdvancedMD, ModMed, Open Dental, Azalea Health, OpenPractice and Epic are named as well. No customer is named running a two way connection in production.

BB on Deployment Model and Data ResidencyOptions and residency are stated with isolation or the processing path left open.
Vendor Published

All 16 subprocessors on the trust center list, from Google Cloud Platform for hosting to Cerebrium and Baseten for model inference and Telnyx and Jambonz for messaging and telephony, are located in the United States. Enterprise customers can ask for tenant isolation, stricter hosting requirements, more isolated deployment patterns and configurable onshore or offshore workflows, and SuperBill's service terms allow both onshore and offshore labor on the service.

The practice chooses whether work arrives by CSV, API or a direct connector. Where recordings and transcripts rest by default, which isolation option comes standard, and where offshore staff sit when a workflow uses them are left to each arrangement.

Commercial
CC on Commercial TransparencyNo price is published and the posture is discoverable: a buyer can establish how the product is sold and what drives the cost before contacting the vendor. Most of the index sits here.
Vendor Published

No price is shown for the agents, the Requests API or Agent-to-Agent RCM. New customers start with a pilot on their own payer mix, which SuperDial suggests running for 30 to 90 days to validate the return, and a new practice system integration, with the calling agent, triggers and write back, can be trialed free. SuperBill's service terms set charges through applicable site terms or a signed order form.

Earlier case studies put the value per call, while Agent-to-Agent RCM measures success by resolved requests, so the unit a customer pays on, whether per attempt, per resolved request, per minute or committed volume, comes with the quote.

BB on Setting and Specialty CoverageCoverage is named with validation behind part of it.
Vendor Published

Revenue cycle outsourcers and billing companies are the core buyers, with United Medical Monitoring, MBW RCM and CBS Medical Billing among those named, alongside specialty groups and dental organizations. Apex IONM and United Medical Monitoring use it for surgical neuromonitoring claims, Mentaya for behavioral health verification and claims, and West Coast Dental in a dental revenue cycle.

Omega Healthcare, Asembia, Henry Schein One, InFocus Eyecare, Guardian and Riverside are among the organizations shown on the site, and the pitch extends to health systems, digital health platforms and payer operations. The workflows covered are benefits verification, prior authorization, claim status, credentialing and enrollment, and provider data.

Alternatives

Alternatives to SuperDial

Other products that do the same kind of work as SuperDial, and what each one does.

Innobot Health automates revenue cycle work for hospitals, health systems, provider groups and RCM and billing companies in the United States.

Spike Technologies sells Health OPS, voice AI agents that take over phone work for physical therapy, occupational therapy, speech language pathology and pediatric therapy clinics.

Needletail AI verifies dental insurance eligibility and benefits for group practices and DSOs and writes the results into the practice management system before the patient arrives.

InsideDesk sells DSOs and multi location dental groups a revenue cycle platform their central billing teams run. InsideAssist syncs open claims from the practice system and payer portals each day, retrieves EOBs and surfaces the next action on each claim.

Voice AI agent platform for the calls that stand between a patient and a prescribed treatment.

Patient access and revenue cycle automation that combines AI agents, rules-based automation agents and a staff of more than 6,000 onshore and offshore specialists inside a platform the company calls its Healthcare Revenue Operating System.

Commercial

Pricing

Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.

Entry Price Pricing Basis BAA Tier Implementation Source
Contact sales
SuperDial quotes each customer after a pilot on its own payer mix, which it suggests running for 30 to 90 days. SuperBill's terms set charges through a signed order form, and whether a customer pays per attempt, per resolved request or by committed volume is settled in the quote. Every customer signs a business associate agreement before protected health information is shared. The agreement text and its subcontractor terms are not public. A new practice system integration can be trialed free on the customer's own workflow. No production setup or customization fee is stated. Vendor Published

No price is shown for the voice agents, the Requests API or Agent-to-Agent RCM. A pilot can include the calling agent, practice system triggers and results written back to patient or claim files, and work can also arrive by work queue export or API.