Electronic Data Capture, built like modern software

Commercial EDC platforms are expensive, closed, and dated. The open-source alternatives run on legacy stacks. edc-core is a from-scratch EDC built on modern web technology, an analytics-native data architecture, and CDISC standards, designed so that regulatory expectations (21 CFR Part 11, ICH E6(R3)) are structural properties of the system rather than afterthoughts.

Get started Five-minute tour GitHub

The analytics workbench: a SQL editor over an immutable study snapshot, with a results grid — The analytics workbench: self-service SQL over an immutable, version-pinned study snapshot.

Why edc-core

No vendor lock-in

AGPL-3.0 licensed: nobody can take this code and sell it back to you as a closed platform.

Programmable study builds

First-class CDISC ODM v2.0 import/export means study builds can be file-driven, code-driven, and LLM-assisted. The visual builder writes through exactly the same versioned-metadata API.

Self-service analytics, in the EDC

Data managers query published study snapshots with real SQL, R, and Python, not a bespoke, vendor-locked reporting layer. Every execution is read-only, sandboxed, version-pinned, and audited. And a listing’s rows can become site queries in bulk, so the workbench doubles as the study’s data-cleaning surface.

Compliance by construction

Append-only audit trails enforced at the database level, Part 11 e-signatures bound to record versions, an audit-trail review UI, and a validation pack shipped with every release.

Standards as the interface

ODM v2.0 (XML + JSON) for metadata, Dataset-JSON v1.1 for data exchange, CDASH-aligned demo CRFs, and a self-contained study archive that outlives the running system.

Boring to operate

The supported production shape is one compose file on a single VM: pinned release images, automatic TLS, paired database-and-lake backups, and cloud-init or Terraform recipes for the major clouds. See Deployment.

Where to start

Installation

Run the full stack locally with Podman or Docker, or deploy the published container images.

Five-minute tour

Seed the demo study and walk the whole lifecycle: entry → query → verify → sign → snapshot → analyze.

User guide

Study builds, data capture, review workflows, and the analytics workbench.

Compliance

How edc-core maps to 21 CFR Part 11 and ICH E6(R3), and what the validation pack contains.

Start here by role

New to EDC

Entry-level data managers: vocabulary first, then a guided loop through entry, review, coding, and analytics.

Study builders

Experienced data managers: versioned builds, protocol import, site layouts, and integrations.

CRO and sponsor leadership

Compliance posture, oversight evidence, and the exit path, without touching a form.

Clinical operations

Site-facing workflows: subjects, entry, queries, and signatures.

Status: alpha. The full workflow (protocol-first or forms-first build, capture, clean, code, sign, snapshot, analyze with SQL, R, and Python, export) is released as v0.5.1, but edc-core has not yet been used in a production study. Follow progress on GitHub.