Talk to us

Ways to work with us

Three ways to work with us, plus custom work. Each one lists what it does today and its limits. Prices are agreed per pilot.

Test data built to your test planAvailable

You list the situations your software must handle, for example “60 patients on metformin whose latest eGFR is below 30”. We build a synthetic FHIR R4 population that contains them in the numbers you ask for. Separate code that does not know how the data was built verifies it.

You receive FHIR R4 data, one test file per patient, a coverage report, a Data Passport, and a manifest that rebuilds the identical dataset.

Limits todayThe library covers medication safety and data integrity; we add situations for your domain. Clinical situations not yet signed off by a clinician are delivered labelled as such. Standard drug and lab codes in the data are in development.

Ask about this

Gap-fill your existing test suiteAvailable

You share your existing FHIR test files. We show which of your situations they already contain and which they lack.

We build only the missing patients and merge them in. Your files are not changed. You get a complete, verified suite and a Test Passport.

Limits todayYour suite must be FHIR R4 JSON. Where it uses medical codes we do not map yet (such as LOINC or SNOMED), we report what we find as a minimum rather than an exact gap.

Ask about this

Failure to regressionBeta

We run your system against the test patients and judge each answer. When your system fails, we pinpoint the smallest change to the record that causes the failure, and save it as a regression case that rebuilds exactly.

After your fix, we rerun it and report whether the failure is fixed, still reproduced, or has changed.

Limits todayProven against an open-source FHIR server and a reference system, not yet a customer’s. Your system must answer over HTTP in JSON. Offered as a guided engagement.

Ask about this

Custom scenarios and data-quality setsAvailable

Custom scenarios for your domain.

Data-quality and robustness test sets: missing values, duplicates, conflicting records, broken references, each one declared and reported.

Limits todayNew clinical situations need clinician review before release. Every defect is declared and reported.

Ask about this

How a pilot works

From a first call to test data you can load. The last two steps are optional.

  1. 1

    A first call

    What your software does, and how you test it today.

  2. 2

    Your situations

    Send 5 to 10 patient situations your software must handle. A non-confidential list is enough.

  3. 3

    Specification

    We turn them into a test specification for you to confirm.

  4. 4

    Delivery

    Test data, test files, coverage report and Data Passport.

  5. 5

    Optional

    We gap-fill your existing suite.

  6. 6

    Optional (beta)

    We run your system and return any failures as regression cases.

Not offered

So that nothing on this site implies otherwise.

  • A web app, dashboard, hosted service or public API
  • Processing or synthesising from your real patient data
  • An anonymisation service
  • A downloadable, installable or containerised product
  • Certification, compliance sign-off, or clinical validation of anything
  • Integration with Epic, Oracle Health/Cerner or any named EHR

Who it is for

Do you have data you cannot use, or no data at all? We work with synthetic data only: we do not take in real patient records.

FirstHealthcare AI and clinical AI companiesEvaluating agents that read and write patient records; known-answer test populations; regression on new model versions.
SecondDigital health, EHR and clinical workflow softwareFHIR development, integration testing, medication reconciliation, release regression.
ThirdHealth-data and interoperability infrastructureMalformed data, broken references, duplicates, event ordering.
LaterPharma and life-sciences technology teamsDevelopment and test environments where real patient data cannot be used.

What teams ask us for

What we can do today, what is in beta, and what is in development.

  • “Give me 5,000 synthetic FHIR patients containing these exact situations.”Available
  • “Give me deliberately incomplete, contradictory or malformed records.”Available
  • “Reproduce the same healthcare cases on every release.”Available
  • “Show me which patient situations my current test suite does not cover.”Available
  • “Test my agent on the missing situations and save every failure as a regression case.”Beta
  • “Give me patients where what is true differs from what the EHR says.”In development
  • “Create the same patient twice, with exactly one workflow error in one of them.”In development
  • “Generate a handoff failure where the receiving system is missing a critical fact.”In development

What we ask in a first conversation

  1. How do you create difficult patient test cases today?
  2. How do you know your current test patients cover the situations your system must handle?
  3. What is the last patient situation that was difficult to reproduce?
  4. How do you recreate a production failure for regression testing?
  5. Can you give us one non-confidential scenario your current test data struggles to cover?

Bring the list of situations your software must handle.

We will turn it into a test specification and show you what a pilot would deliver. Already have FHIR test data? Send a sample and we will show you what it is missing. Or email us at [email protected].

Talk to us about a pilot