Structured + unstructured
One platform for tables and free text: clinical notes, transcripts, documents, synthetic data.
EviData™ by Woodway Assurance
The privacy layer for the AI data economy.
EviData measures re-identification risk, guides anonymization with AI agents, and delivers auditable evidence that your data meets recognized privacy standards — structured or unstructured.
Runs entirely in your environment. Your data never leaves.
trusted by data leaders
- Archimedes
- Environics Analytics
- BORN Ontario
- Nina Capital
- Weeks → Minutes
- consultant assessments, automated
- 3
- data classes: de-identified, anonymized, synthetic
- ISO/IEC 27559
- standards-aligned assessment
- 100%
- in your environment — data never leaves
Some figures illustrative for design review — pending verification.
the problem
Your most valuable data is stuck.
01
AI teams need it. Privacy teams can't sign off on it.
02
Manual expert reviews take weeks and don't scale.
03
Unstructured text — notes, transcripts, documents — goes unused entirely.
04
“Probably fine” isn't evidence a regulator will accept.
why evidata
Three advantages. One platform.
Fully automated risk assessment
Quantified re-identification risk against recognized thresholds. Measured, not guessed — in minutes.
AI agents that explain, plan, and guide
EviAgent™ interprets your results, plans the de-identification strategy, and executes privacy-preserving transformations — like an anonymization expert on demand.
the agent team
A multi-agent system built for the de-identification challenge.
EviData 3.0 is a multi-agentic AI platform. Each agent owns one step of the assurance loop — together they take a dataset from unknown risk to defensible evidence.
agent/assessor
Assessor
Measures re-identification risk quantitatively, against recognized thresholds.
agent/explainer
Explainer
Turns every finding into plain language your privacy team — and your regulator — can follow.
agent/planner
Planner
Designs the de-identification strategy that clears thresholds while preserving utility.
agent/transformer
Transformer
Executes privacy-preserving transformations in the same workflow. No handoffs.
EviAgent™ — live session
Can we release this oncology dataset to our lab customer?
EviAgent
Re-identification risk 0.14 — above the 0.09 threshold. Drivers: rare diagnosis codes × 5-digit ZIP.
EviAgent
Plan: generalize ZIP to 3 digits, top-code age 90+. Projected risk 0.06 ✓ — utility preserved 96% [projected].
the assurance loop
From unknown risk to defensible yes.
- 01
Connect
Bring your dataset. It never leaves your environment.
- 02
Assess
Automated re-identification risk measurement, scored against recognized standards.
- 03
Understand
EviAgent explains every risk in plain language.
- 04
Reduce
AI-guided anonymization preserves maximum data utility.
- 05
Prove
Download an auditable, standards-aligned report. A defensible yes.
who it's for
Built for both sides of the data economy.
ai data intermediaries
You sell trust in data. Now prove it.
Certify training data as de-identified before it reaches your lab customers — privacy assurance as a product feature.
healthcare & life sciences
Unlock clinical data.
For research and AI under HIPAA, GDPR, and provincial guidance.
enterprise data teams
Share and license with evidence.
Analyze and monetize sensitive data with proof it's safe.
privacy officers & counsel
Reports built for scrutiny.
Standards-aligned, auditable evidence — regulator-ready.
client examples
Proof in production.
health data platform
Archimedes
Independent assurance for data made available to researchers.
data & analytics
Environics Analytics
Verified de-identification for commercial data products.
provincial registry
BORN Ontario
Defensible secondary use of sensitive health data.
Vignette details condensed from public announcements — to be confirmed with each client.
compliance & trust
Standards-aligned. Built to be examined.
- ISO/IEC 27559:2022
- GDPR & Quebec Law 25 anonymization
- Ontario IPC guidelines
- HIPAA Expert Determination methodology
Founded by Dr. Khaled El Emam, Canada Research Chair in Medical AI and one of the world's most cited experts in anonymization.
Are you confident your data can stand up to scrutiny?
Get an independent answer in minutes — not weeks.