OpenAdapt
Start your first workflow

← All workflow templates

Workflow templateProven reference

Automate patient record creation in OpenEMR

A complete synthetic patient is created from structured demographics on a pinned local OpenEMR 8.0.0.3 fixture. The run returns VERIFIED only when a separately authenticated REST readback agrees with a direct SQL read and a non-target table-delta audit - the screen never certifies its own write.

This workflow runs today, end to end, against the named open-source reference application in the openadapt-flow repository. The steps below are the real demonstrated steps and the verification is the real oracle.

Published results — exactly as measured

Trials
3
Verified / expected halts
3 / 0
Silent incorrect successes
0
Model calls per run
0
Duration
59.8 s median

published median end-to-end runtime across the three Standard-profile runs. Source: Reference qualification published on openadapt.ai and app.openadapt.ai/demo. Measured on compiler 1.23.0 line; evidence pack fe6a3e778f16 referenced by the live demo.

Provenance

Footage and evidence media

Runs on

OpenEMR 8.0.0.3 on a pinned local synthetic fixture - fake patients only, everything binds to localhost. Never point this at a real install without qualification.

Source in openadapt-flow →

The demonstrated steps

  1. 1Log in to the OpenEMR fixture as the demo admin
  2. 2Open the patient creation flow from the main dashboard
  3. 3Enter the structured demographic fields - each value a parameter substituted per run
  4. 4Save exactly one complete patient record
  5. 5Wait for the independent effect check: separately authenticated REST readback must agree with direct SQL and the non-target delta audit

Parameters: the structured demographics demonstrated in the recording (one parameter each) — recorded values are the defaults; every replay can override them.

How the outcome is verified

All three fresh Standard-profile runs created exactly one synthetic patient and returned VERIFIED only after a separately authenticated REST readback agreed with a direct SQL read and a non-target delta audit. Observed silent incorrect success was 0/3.

  • separately authenticated REST readback
  • direct SQL read of the patient row
  • non-target table-delta audit (exactly one new patient, nothing else changed)

Questions about this template

How is success established for this workflow?

By an out-of-band check of the system of record: a separately authenticated REST readback that agrees with a direct SQL read plus an audit that no other table changed. The browser session cannot certify its own result.

Source: How the result was checked (landing page)

What did the three published trials cost in model calls?

Zero. All three Standard-profile runs recorded zero model calls, and observed silent incorrect success was 0 out of 3.

Source: Reference qualification stats

Can I run this against a real EMR?

Not without qualification. The published evidence is against a pinned local synthetic fixture with fake patients; a real deployment requires your own identity, effect, and policy contracts qualified against your system of record.

Source: Qualification approach

Try it from the command line

The compiler and runtime are open source and MIT licensed. Healthy runs are local and make no model calls.

pip install 'openadapt[browser]'

Installs the OpenAdapt launcher, compiler, and browser capability.

openadapt flow record --url http://localhost/openemr --out rec && openadapt flow compile rec --out bundle --name create-patient

Demonstrate or reuse the recording, compile it into a deterministic bundle, then replay locally.

Put this workflow into production

Bring your version of this workflow and the record that proves its outcome. We'll map the deployment, verification, shadow run, and supervised rollout.

Evaluate a workflowRead docs