A research-grade synthetic data engine wrapped in a simple web app.
Every patient gets a plausible life history — bimodal age distribution, sex, pregnancies and hospital admissions — generated from configurable population statistics.
Comorbidity risk models for HIV, acute kidney injury → chronic kidney disease and type 2 diabetes, with population parameters you can tune before each run.
Age- and sex-aware creatinine trajectories anchored to 800+ real CALIPER reference-range datasets, perturbed with realistic biological variation.
Each patient is labelled on every birthday and around each admission, capturing the modeled lab value and the conditions active on that exact date.
One-click HL7 v2.5.1 ORU^R01 output so you can drive a real Laboratory Information System with lifelike results — without touching real patient data.
Saved runs store their parameters and random seed, so any run can be repeated later and reproduces byte-identical output.
From sign-in to a downloadable dataset in five steps.
Sign in with Google (or the local dev login). Your generation quotas and saved runs are tied to your account.
Set the number of patients to generate. Optionally open "Epidemiology parameters" to adjust demographics, admission rates, pregnancy and comorbidity risks.
Review the population across the Demographics, Admissions, Comorbidities and Creatinine charts, plus a preview data grid.
Export the CSV bundle (a patients table and a per-date annotations table) or the HL7 message file to feed into your LIS or analysis pipeline.
Open "Saved runs" to repeat any previous run with the exact same parameters and output.
Copyright © fAIkserum 2026. Synthetic data only — not for clinical use.