Clinical Decision Support
D

Diagnosis Pad

Diagnosis Pad is an offline clinical assistant for doctors that listens to a consultation and, as the conversation develops, offers the top three differential diagnoses, the top three recommendations with suggested questions to ask, and a clinical note summarizing the transcript at the end. A clinician can open the reasoning behind a suggested diagnosis, look up any of more than 6,000 conditions, symptoms and procedures in the built in Clinical Encyclopedia, and edit the note. Settings decide which outputs appear live and which wait for the end of the session. The App Store listing, titled AI Doctor, also presents it as a symptom checker.

Transcription and inference run on the user's device, so no consultation is sent to the cloud. The app runs on compatible iPhones and iPads, Apple Silicon Macs and recent high performance Android devices, in English and Arabic. The company cites training on up to 9 trillion tokens, more than 680,000 hours of speech for its recognition engine and a patent pending design, and the app runs on the Inferencer engine. Enterprise installs the models on a customer's own servers with custom electronic health record integration.

Dr. Nora, a family doctor and medical AI researcher, founded the product, which launched in January 2025 and is published on the App Store by Ashraf Samy. The free plan includes differentials, guidance and notes with a time limit per session. Essentials costs $9.99 a month without the limit, and Professional costs $29.99 a month with session history.

AI Health Index verifiedNovember 19, 2026
Compare Diagnosis Pad with other vendors
Founded
2025
Headquarters
—
Categories
clinical-decision-support, ambient-scribes
Indexed Products
Diagnosis Pad, Clinical Encyclopedia, Diagnosis Pad Enterprise
Buyer Segments
Individual clinicians, Medical practices, Healthcare organizations
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

Inference is the product. Speech recognition turns consultation audio into text on the device, and a local model generates the differentials, recommended questions and follow up guidance, then drafts the clinical note and expands the rationale behind any suggestion on request. The Clinical Encyclopedia and its context aware terminology database support that workflow.

Recording controls, session history and sharing organize the work, but there is nothing to use without the model. Every output a clinician sees during a session, whether live or at the end, comes from inference running on the phone, tablet or Mac in the room.

CC on Autonomy and Oversight ModelAutonomy is claimed and oversight is asserted without a mechanism. Human in the loop appears as a phrase rather than a described control.
Vendor Published

A practitioner starts each session and chooses whether differentials, guidance and notes appear live or after recording ends. Each proposed differential shows its rationale, notes can be edited, and the terms tell healthcare professionals to apply their own judgment.

Live suggestions reach the clinician with no approval step, which suits an assistant but matters more when the same app is used as a symptom checker. There is no escalation for urgent symptoms, no review requirement when a patient rather than a clinician uses it, and no threshold that holds back a low confidence result.

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

The app runs on the Inferencer engine with deterministic inference. It transcribes speech locally and updates the differential, suggested questions and note as the consultation adds information. The company cites model training on up to 9 trillion tokens and more than 680,000 hours for the speech engine, a terminology database of more than 6,000 terms and a patent pending design, and the App Store release history logs tuning of explanations, diagnoses and notes.

The language and speech models are not named, nor are their training sources, and no mapping links an app release to the model versions inside it.

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

Consultation audio and text are processed on the clinician's device, and the privacy policy says the app sends no data to online servers for model processing, so no model provider receives patient information in standard use. Enterprise inference runs on the customer's own servers. Data can leave the app only when a user turns on operating system backups, shares a session or submits support material after reviewing it.

The base models are not identified, and neither are any providers involved in an enterprise configuration. Inferencer, the engine the app runs on, is named without saying whose models it loads.

DD on Clinical and Operational EvidenceNo named deployment and no performance claim a reader can check. A figure published with no source sits here rather than higher.
Vendor Published

The company reports lab sessions running past 30 minutes on supported hardware, a measure of runtime rather than accuracy. The founder has shown the app at the RACGP Clinical Update Conference and written about early cases, but those predate later releases and use no reference standard. Training scale, up to 9 trillion tokens and 680,000 hours of speech, describes the inputs rather than the results.

No study measures how often the differentials are complete or correct, how good the suggested questions or notes are, or what the app changes in clinical use for the current release.

BB on AI Safety and PHI StewardshipCategorical commitments are published, such as no training on customer data, without the retention schedule behind them.
Vendor Published

The consultation stays on the device in standard use. Users can set an automatic deletion policy, Professional keeps prior sessions on the device, and Apple or Google backups include local data only if the user turns them on. Feedback carries only what the user chooses to send after reviewing it, and the privacy policy also covers optional anonymous diagnostic logs and searches of outside health websites that the user launches.

No default retention period applies across all tiers, and data handling commitments for support uploads and enterprise deployments are not given. Nothing describes how the developer responds once a misleading diagnostic suggestion comes to light.

Regulatory and Compliance
CC on HIPAA and BAA PostureA status is stated but not supported: compliance is claimed, or business associate status stated, without the underlying document, including where the only privacy notice published covers the website rather than the service that handles patients. The mirror case grades here as well: a substantive privacy document that reaches the service, with no statement of business associate status anywhere and no scope position taken. One of the two elements the band above asks for is present, which is a grade below it and not a grade at the floor.
Vendor Published

The site displays HIPAA, GDPR, NHS, PIPEDA and Australian Privacy Principles compliance marks. The privacy policy says consultation data stays on the device in standard use, apart from optional backups and sharing or feedback the clinician initiates, so the developer sits outside the ordinary path of a consultation. Enterprise is a separate configuration on the customer's servers.

The terms do not say whether the developer takes a business associate role for support or enterprise work, how health information sent voluntarily to support is handled, or whether an agreement is available on any tier.

DD on Security Certifications and Trust CenterControls are asserted with nothing independent behind them, or nothing is published. Read the note before concluding anything: this is the grade most often corrected on a second pass, because assurance material frequently sits on a parent domain or inside an old announcement rather than on the product pages.
Vendor Published

Local inference keeps the standard path for consultation data on the device where the session happens, and users can set automatic deletion and decide whether to share or back up a record. The online functions are limited to subscription checks, operating system crash reports and support submissions.

Compliance marks appear on the site, but no security audit, certification or independent penetration test of the app or its developer is reported, and controls on enterprise support access are not described. In practice, the device's own passcode, encryption and backup settings carry much of the protection.

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 terms say the app gives health professionals additional information and leaves decisions with the practitioner. In use it generates patient specific differentials and suggested questions during a consultation, shows the reasoning behind each, and transcribes the clinician's own encounter, so the basis for a suggestion can be inspected. The store listing, titled AI Doctor, also offers it as a symptom checker.

A general medical disclaimer is the only regulatory statement. No FDA authorization, named device exemption or Australian TGA registration is given, which leaves the symptom checker use, outside a consultation a practitioner leads, the most exposed.

DD on AI Governance and Bias DisclosureNothing published on how model behavior is governed or tested. Multilingual operation with no subgroup performance sits here when the vendor markets recognition quality as a strength, because a caller the system failed to understand leaves no complaint and no record.
Vendor Published

A clinician can expand the reason behind a differential and look up related terms in the offline library, and the company describes its inference as deterministic, so the same input should give the same output. Changes to model behavior, explanations and notes are listed in the App Store release history.

No evaluation compares diagnostic suggestions across ages, sexes, specialties or the app's two languages, and no process decides whether a model update is fit for clinical use before it ships. How often systematic errors occur, and what happens when one is found, is not described.

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

The clinician can check a suggestion against its stated rationale, read the full transcript and edit the note before using it. A feedback function lets a user review and edit material before sending it to the developer, which gives a route to report a bad output without exposing more of the consultation than intended.

The terms disclaim warranties and liability for loss or for inaccuracies in generated content. They set no response time for a report of an incorrect clinical suggestion and no commitment to correct or compensate for a decision based on one, so recourse rests on the clinician's own review.

Integration and Deployment
DD on EHR and Interoperability DepthNo integration evidence. A connector described as available on request grades here until one exists.
Vendor Published

The individual plans have no record connection. After a session, the clinician shares or exports the note through the device's operating system to whatever destination they choose and enters it in the chart separately. Session sharing is included on the free plan and transcript editing on the paid ones, which shapes what text reaches the chart. Enterprise offers custom electronic health record integration with inference on the customer's servers.

No hospital system, interface standard or live enterprise connection is named, so whether an enterprise setup reads from or writes to a chart depends entirely on the custom work.

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

The individual app runs on a compatible iPhone or iPad, a Mac with an Apple M1 chip or later on macOS 14.6, or a recent high performance Android device, and consultation data stays on that device for inference. Enterprise runs the models on the customer's own servers. Optional Apple or Google account backups, sharing through the operating system and searches the user launches are the paths by which data can leave the device.

For enterprise installations, tenant isolation, the location of support data and the terms of administrative access are not described. In the individual product, residency follows the device and its settings.

Commercial
AA on Commercial TransparencyPublished tiers with figures, a stated unit of charge, and a route to start without a sales conversation.
Vendor Published

Pricing is listed in US dollars. The free plan includes on device processing, transcription, differentials, guidance, notes, detailed reasoning and session sharing, with a time limit on each session and no history. Essentials at $9.99 a month removes the limit and adds transcript editing, and Professional at $29.99 a month adds review of past sessions. Subscriptions run through Apple or Google, which convert prices to local currency at checkout.

Enterprise, with on premises hosting and custom record integration, is quoted through a contact form, and its implementation fees are not listed. Hardware requirements, an M1 Mac or a recent high performance phone, decide who can use the individual plans.

CC on Setting and Specialty CoverageCoverage is claimed broadly without specifics, or stated clearly with nothing validating it yet.
Vendor Published

Clinicians with a compatible device who capture the encounter as it happens are the intended users, and the app came from a family physician. The encyclopedia spans conditions, symptoms and procedures, and the app store listings add a symptom checker path for individuals, in English and Arabic.

The evidence is narrower than that breadth. No specialty testing exists for the differentials or notes, and no validation covers settings such as emergency care or telehealth. Hardware limits affect who can run it during a visit, since older phones and Intel Macs are excluded.

Alternatives

Alternatives to Diagnosis Pad

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

Medows is a clinical workspace for doctors on ward rounds and on call shifts, designed alongside junior doctors in Indian hospitals.

Japanese health technology company with two AI products.

Dr.Oracle is a medical question and answer assistant for physicians, residents, physician assistants, nurse practitioners and medical students.

AI4Eyes is building OSDx, a compact imaging device that mounts on a clinic's slit lamp and consolidates ten dry eye tests into one five minute scan, paired with subscription AI software that returns a dry eye disease diagnosis and treatment recommendation for the clinician to approve.

Corti is a Copenhagen company that builds healthcare AI models and sells them as infrastructure.

Pymander Health is a free AI doctor that adults in the United States reach by text message, over SMS or iMessage, with no app to install and no account needed for a first consult.

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
$0 with session limits
$0 baseline
The free plan includes local transcription, differentials, guidance and notes with a session time limit. Essentials is $9.99 per month and removes that limit. Professional is $29.99 per month and adds access to prior sessions. Enterprise is quoted. The standard app processes consultation data on the device, with device backups and user initiated sharing optional. No business associate agreement position is stated for the individual or enterprise plans. The individual plans are available directly through Apple or Google with no setup. Enterprise offers installation on the customer's servers and custom electronic health record integration, with charges set in the quote. Vendor Published

Apple or Google converts the prices to the user’s currency at checkout. The paid individual tiers support transcript editing. Enterprise contract terms vary with hosting and integration scope.