PRACTICE LABS

Learn by doing — in working labs.

Hands-on environments with synthetic patients you can't break. Every lab runs at your level: a guided visual mode for beginners, a pro mode when you're ready.

All data is synthetic · nothing real, ever30 labs Free cheat sheets

Every lab runs in two modes — what do Guided and Pro mean?

The tags on each card show which modes that lab offers. You pick your mode once in settings and every lab follows it.

GuidedIncluded free

The lab walks you through it. Starter code is written for you, steps are broken down, and a visual mode replaces anything you would otherwise type from memory. Built for a first encounter with the format or the tool.

ProPaid plan

The scaffolding comes off. You get a blank editor, the real tooling professionals use, and the advanced material Guided holds back — such as the full billable code set in the terminology labs. Built for practising the way the job actually works.

Not sure? Start in Guided. You can switch whenever you like and your progress carries across.

Category

Track

Format

Access

30/30

Digital Health & Informatics

28 labs

The core track: how modern healthcare moves, stores and reasons over information — standards, terminologies, integration, clinical decision support, and the data & research methods behind evidence-based care. Leads to the SIDHI certification.

Data Foundations — start here

Data Formats Playground

No jargon, no code, no login. See one blood-pressure reading written as plain English, then as CSV, JSON and XML — then check your own files in a live inspector that explains every error in ordinary words.

JSONXMLCSVBeginnerNo code

GuidedPro

Format Converter Bench

Watch a single patient record turn into CSV, JSON, XML, HL7 v2 and FHIR side by side, with the same field lit up in all five. Then paste your own data and convert it between formats.

CSVJSONXMLHL7 v2FHIRConversion

GuidedPro

API Playground

How one piece of software asks another for something — explained with no code and no login. Dissect a real request piece by piece, then drive a live practice API that makes you meet 401s, rate limits, pagination and duplicate-payment bugs on purpose.

APIHTTPRESTOAuth2SMARTWebhooksBeginner

GuidedPro

Version Control Playground

Git and GitHub explained with no command line and no login — repo, commit, branch, pull request, merge, deploy. Follow one change to a clinical rule from edit to live, and learn what to do when someone commits an API key or patient data by mistake.

GitGitHubPull requestsCI/CDBeginnerNo code

GuidedPro

Interoperability & Integration

FHIR Mapper · HL7 v2 → FHIR R4

Map a hospital admit message into a FHIR Patient resource, field by field, and validate it against a profile. Visual mapper or pro editors.

HL7 v2FHIR R4FHIRPath

GuidedPro

FHIR API Sandbox

Make live FHIR REST calls against a synthetic server — GET a Patient, POST an Observation, run a search, create a Bundle transaction. See the request and response in full.

FHIR R4REST APIResources

GuidedPro

FHIR Conformance & Profiling

The rigorous 'structured FHIR' skill employers test: how a StructureDefinition constrains a base resource into a profile (US Core, India's NDHM IG), binding strength, must-support, slicing, and whether an instance actually conforms. Eight graded conformance scenarios.

FHIR R4ProfilingStructureDefinitionUS CoreConformanceNDHM IG

GuidedPro

HL7 v2 Message Explainer

Paste any raw HL7 v2 message and get a plain-English, segment-by-segment breakdown of every field — the fast way to learn the format.

Guided

Clinical Integration Engine

Trace a message through a hospital integration engine — routing rules, transformation steps, ACK/NACK handling, and multi-destination delivery. The plumbing that connects hospital systems.

HL7 v2FHIR R4RoutingACK/NACK

GuidedPro

CDS Hooks · Decision Support

Author a CDS service response — cards, suggestions, and feedback links — triggered by real EHR hook events. Understand how decision support fires at the point of care.

CDS HooksFHIR R4Clinical Decision Support

GuidedPro

Clinical Systems & Workflow

EHR Workflow · CPOE & Order Safety

The clinician-facing layer: place orders on a synthetic patient and watch a real check engine fire allergy, drug–drug interaction, renal and duplicate-therapy alerts as you type. Plus order sets and the workflow judgement — wrong-patient entry, alert fatigue, and closing the administration loop.

CPOEe-PrescribingOrder setsPatient safetyEHR

GuidedPro

Nursing Informatics · Care Plans, eMAR & NEWS2

The administration end of the medication loop, where CPOE hands over. Build a care plan from an assessment, run barcode medication administration against a real five-rights engine on a ward where two patients share a surname, and score a quietly deteriorating patient with the real NEWS2 table.

Care planseMARBCMANEWS2Patient safetyADPIE

GuidedPro

Simulation & Digital Twins

Digital Twin · Physiological Simulation

A real glucose–insulin model running in your browser. Calibrate insulin sensitivity until the twin reproduces this synthetic patient's actual CGM curve — then ask it what-if questions. The difference between a generic simulation and a twin is the calibration.

Digital TwinSimulationODE modelCalibrationGlucose–insulin

GuidedPro

Imaging Informatics · DICOM

Imaging Informatics · DICOM & PACS

The imaging half of health IT: decode a DICOM header tag by tag, build DICOMweb (QIDO/WADO/STOW) requests, trace the Patient›Study›Series›Instance model, and walk the radiology workflow from order to report. Synthetic data, no PHI.

DICOMDICOMwebPACSModality WorklistRadiology

GuidedPro

India Health Stack · ABDM

ABDM · India Health Stack

Work the consent flow that powers India's health data exchange — HIP, HIU, Consent Manager and Gateway — plus ABHA identity, the ABDM building blocks, and the NDHM FHIR Implementation Guide. The India-native skill no global platform teaches.

ABDMABHAConsent (HIP/HIU)NDHM FHIR IGIndia

GuidedPro

Terminology & Coding

Shabda · Terminology & Code Lookup

Look up any clinical concept — SNOMED CT, ICD-10-CM, ICD-9-CM, LOINC — and trace its crosswalks, hierarchy and coding rules. Switch to Pro mode to reveal all billable codes.

SNOMED CTICD-10-CMLOINCCrosswalk

GuidedPro

ICD-10-CM Clinical Coding

Decode the anatomy of real ICD-10-CM codes (category, laterality, 7th character, placeholder X), then code synthetic cases — the engine validates billability and specificity, not multiple choice.

ICD-10-CMClinical CodingCase Studies

GuidedPro

ICD-11 Post-coordination

ICD-10 is a book that was digitised; ICD-11 is a web service. Learn the skill that cannot be taught from an index — composing a code from a stem plus extension codes, and reading a cluster string the way a terminology server does.

ICD-11Post-coordinationWHOTerminology

GuidedPro

Clinical NLP & AI

Clinical NLP · Annotation & Extraction

Read a synthetic clinical note and label the medicines, doses, problems and lab values — build the gold standard, then review what the AI extracts.

SNOMED CTRxNormLOINCNER

GuidedPro

Clinical AI/ML · Evaluate & Govern

Prove you can judge a clinical model, not just run one: move a decision threshold and watch the confusion matrix, ROC/AUROC and calibration respond; audit fairness across subgroups; write a model card; and rule on EU AI Act, drift and human-oversight scenarios. Synthetic cohort.

ML EvaluationAUROCCalibrationFairnessModel CardEU AI Act

GuidedPro

Revenue Cycle & Medical Coding

RCM Workbench

The all-in-one revenue-cycle environment: scrub claims before submission, work denials by CARC/RARC reason code, and decode X12 835 remittances — with coding-QA and eligibility tools landing next. One robust RCM technology lab, not a single trick.

RCMClaim scrubbingDenialsX12 835Coding QA

GuidedPro

Security, Audit & Compliance

Audit & Compliance Workbench

The governance skill set of the AI-native era: review a synthetic ePHI access log for violations, rule on HIPAA/DPDP compliance scenarios, and audit an AI agent's decisions, plus a breach-notification calculator and a Safe Harbor de-identification tool. Five tools, one workbench. Synthetic data, no real PHI.

HIPAADPDPAudit logAI governance

GuidedPro

Data Science & Research

Health Data Analytics Lab

Write real SQL against a synthetic patient dataset. Query across 5 tables — patients, encounters, diagnoses, vitals, and lab results — to answer clinical questions. Guided mode scaffolds the query; Pro mode is a blank editor.

SQLHealth DataICD-10Cohort Analysis

GuidedPro

Health Data Science Lab

Write Python with pandas and matplotlib against a synthetic 20-patient hospital dataset — all running in your browser via WebAssembly. Five scenarios from demographics to length-of-stay analysis. Guided mode provides starter code; Pro mode is a blank slate.

PythonpandasmatplotlibWASMHealth Analytics

GuidedPro

Biostatistics Lab

Analyse a synthetic randomized controlled trial with Python, scipy and statsmodels — running in your browser. Five scenarios: descriptive statistics, the independent t-test, chi-square, correlation & linear regression, and Kaplan-Meier survival analysis. Guided mode provides starter code; Pro mode is a blank slate.

Pythonscipystatsmodelst-testSurvival

GuidedPro

Epidemiology & EBM Lab

Turn study data into evidence with Python — running in your browser. Five scenarios: disease frequency, measures of association (RR / OR / NNT), diagnostic-test evaluation (sensitivity, specificity, ROC/AUC), confounding & Mantel-Haenszel adjustment, and an evidence-based-medicine appraisal (ARR / RRR / NNT with confidence intervals).

PythonEpidemiologyRR/OR/NNTDiagnostic testsEBM

GuidedPro

Data Harmonization · OMOP CDM

Harmonize messy source data into the OMOP Common Data Model — the foundation of real-world-evidence and pharma research. Map local codes to standard concepts, build a patient cohort, run drug–condition queries, and compute prevalence across standard concepts. Synthetic data only; the CDM structure is open.

PythonOMOP CDMRWEHarmonizationPharma

GuidedPro

Pharmacovigilance · MedDRA

Detect drug-safety signals with Python on a synthetic spontaneous-reporting database. Navigate the MedDRA hierarchy (PT → SOC), code and roll up adverse events, and compute disproportionality (PRR / ROR) with the classic signal criteria. Synthetic MedDRA codes only — the dictionary is licensed.

PythonMedDRASignal detectionPRR/RORPharma

GuidedPro

Life & Biomedical Sciences

2 labs

An adjacent track for computational skills in the life and biomedical sciences — sequence analysis and biomedical signal processing. For biotech, life-sciences and biomedical-engineering learners.

Bioinformatics Lab

Work with DNA and protein sequences in Python — GC content, reverse complement, transcription and translation, FASTA parsing and pairwise alignment — entirely in the browser. Built for biotech, life-sciences and bioinformatics students.

PythonDNA/RNAAlignmentFASTATranslation

GuidedPro

Bioinstrumentation Lab

Process biomedical signals in Python with scipy.signal: generate and filter a synthetic ECG, detect heartbeats, and analyse the frequency spectrum with the FFT. Built for biomedical-engineering students.

Pythonscipy.signalECGFilteringFFT

GuidedPro