E-Discovery Platform

Part of the worlds-biggest-software-project initiative.

An open-source, AI-native e-discovery platform built to the EDRM framework — covering legal hold, collection, processing, review, privilege logging, and production.

E-Discovery Platform is a community-built alternative to the proprietary tools (Relativity, Everlaw, DISCO, Exterro) that dominate litigation technology. It targets government agencies, law schools, legal aid organisations, and small-to-mid-size firms that cannot justify $20K–$500K/year commercial platforms, while bringing modern AI capabilities — privilege review, proportionality analytics, and custodian interview automation — to the EDRM workflow.

Why E-Discovery Platform?

No substantive open-source option exists. Apache Tika provides document extraction, but no community project has ever built out the full EDRM workflow despite EDRM being an open, freely implementable framework.
Commercial pricing locks out most of the legal market. Per-GB all-in costs run $48–$180/GB; enterprise platforms reach $50K–$500K+/year. Mid-size firms, legal aid, and government bodies are systematically priced out.
Privilege review is the dominant cost driver. Privilege determinations consume 60–70% of total review cost. Incumbents assist with relevance review but do not generate FRE 502-compliant privilege logs end-to-end.
Proportionality analysis is unsupported end-to-end. FRCP Rule 26 requires proportionality assessment before discovery commences, yet no current platform produces a defensible proportionality memo from the data universe.
Cloud lock-in raises sovereignty concerns. Cloud-only vendors (Everlaw, DISCO) cannot serve customers with on-premises or sovereign-data requirements; on-prem incumbents (Relativity, Nuix) carry steep licensing and administration costs.

Key Features

EDRM Workflow Coverage

Document ingestion pipeline using Apache Tika for extraction across 1,000+ formats
Automated deduplication (exact and near-duplicate), OCR, and metadata preservation
Full-text search with boolean, proximity, date, and metadata filtering
Production in court-accepted formats (TIFF + OCR, native, PDF) with Concordance DAT and Opticon OPT load files
Complete audit trail of review decisions for defensibility under FRCP Rules 26, 34, and 37

Legal Hold & Custodian Management

Notice creation, distribution, acknowledgment tracking, and automated escalation
Custodian interview workflow that maps relevant data systems before collection
Integration points for IT asset inventory and enterprise GRC systems
DSAR (Data Subject Access Request) integration for organisations managing both litigation and privacy obligations

Document Review Workspace

Configurable coding fields, batch assignment, and concurrent multi-reviewer access
Email thread visualisation and near-duplicate clustering to reduce redundant review
Role-based access control across reviewer, supervising attorney, case manager, and admin roles
Privilege log generation with standard fields (Bates range, author, recipient, date, privilege basis)

AI-Assisted Review

TAR 2.0 / Continuous Active Learning for document prioritisation and review scope reduction, based on the public-domain CAL algorithm (Cormack and Grossman, 2014)
AI privilege pre-flagging using ML classification of attorney-client and work product communications
Early case assessment dashboard with document universe statistics, custodian volumes, and projected review burden
Cross-matter learning with confidentiality partitioning (backlog)

Defensibility & Compliance

Workflow aligned to the EDRM stages (Information Governance → Production)
ISO/IEC 27050 alignment for international e-discovery processes
FRE 502-aware privilege workflow with attorney-in-the-loop confirmation
GDPR Article 17 / data minimisation handling for EU-sourced data

AI-Native Advantage

Incumbents bolted AI onto legacy review platforms; this project designs around it. AI-native privilege review generates FRE 502-compliant privilege logs with Bates references, author/recipient mapping, privilege basis, and redaction recommendations — a workflow no current platform handles end-to-end. A proportionality analytics module ingests the data universe and produces a defensible FRCP Rule 26 memo. Custodian interview automation conducts structured interviews, maps data systems, and issues tailored hold notices, compressing a multi-week process to days. All AI outputs remain attorney-supervised in line with ABA Formal Opinion 512 and state bar guidance.

Tech Stack & Deployment

Document extraction: Apache Tika (Apache Licence 2.0) as the processing foundation
Search: Elasticsearch / Apache Solr / Lucene compatible
Standards alignment: EDRM framework, ISO/IEC 27050, FRCP Rules 26/34/37, FRE 502
Production formats: TIFF + OCR, native, PDF with Concordance and Opticon load files
Deployment modes: self-hosted (on-prem and private cloud) and managed cloud, addressing the data sovereignty gap left by cloud-only incumbents
Integrations: Office 365, Google Workspace, Slack, Zoom for defensible cloud collection; SFTP and REST API for custom data sources

Market Context

The global e-discovery market is $18.73B in 2025, projected to reach $46.06B by 2034 at a 10.49% CAGR (Fortune Business Insights); cloud accounts for 76.65% of share in 2026. Commercial pricing ranges from $250/month (Logikcull) through $2K–$10K/month (Everlaw, DISCO) to $20K–$500K+/year (Relativity, Nuix, Exterro), with industry per-GB all-in costs of $48–$180. Primary buyers span AmLaw 200 litigation partners, Fortune 1000 in-house legal, mid-size law firms, government agencies (DOJ, SEC, state AGs), and solo/boutique litigators — with the lower-budget cohort largely unserved by current options.

Project Status

This project is in the research and specification phase.
Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. See discussion for context.