// Receiving the data is not the same as using it

Watch fragments become one record.

Fragmented records from different sources do not speak one language until they are normalized and merged at the patient level. Paste two or more records for the same patient, in different formats, and watch the engine normalize the codes, resolve the identity, and merge them into a single record with full lineage.

Runs entirely in your browser. Nothing is sent anywhere or stored.

01 Sources Two or more records for the same patient, from different systems. Format is detected as you type.

Try this

02 Pipeline Intake, normalize, merge. The same three stages as the production runtime.

INTAKE

parse each source format

idle

→

NORMALIZE

one schema, one vocabulary

idle

→

MERGE

resolve to one patient

idle

01Intake

02Normalizecodes and fields mapped to one common model

03Mergeone patient, every value accountable to its source

Agreed — sources match Reconciled — conflict resolved Union — added by one source Single source

What this demo does and does not do

It genuinely handles

Common HL7 v2 segments (PID, DG1, OBX, AL1, and visit/header segments), FHIR R4 Patient, Condition, Observation, MedicationStatement, and AllergyIntolerance, and a claims CSV row.
A real two-way code crosswalk across ICD-10, SNOMED CT, LOINC, and RxNorm for common concepts, plus date, name, sex, and unit normalization.
Transparent identity resolution and conflict resolution, with the rule shown for every decision.

It does not pretend to

Parse C-CDA / XML or raw X12 EDI, or cover every segment and resource.
Replace the full terminology graph. Codes outside the curated set are preserved and labeled "unmapped," never invented.
Be a production master patient index. The match score is a transparent heuristic. Data is synthetic; nothing is stored.