Care with PHI. Code without it.
Clinical and operational workflows run on governed access to PHI — tokenized datasets for analytics and research, full audit trails for reviews.
Why this buyer hurts.
PHI everywhere
Claims, scheduling, and support flows scatter health data across vendors and warehouses.
Research friction
De-identifying datasets for analytics takes weeks of one-off engineering per study.
Audit exposure
Access reviews fail when nobody can prove who saw which record, when, and why.
Governed PHI by default.
cloakit isolates PHI in the vault, serves each workflow only what its purpose allows — full, masked, or redacted — and records every access with actor and purpose.
Pattern: purpose-scoped detokenization, tokenized research extracts with redaction, and review-ready audit exports.
The workflow, end to end.
Classify the PHI
Tag fields by sensitivity and workflow: care delivery, billing, or research. Policies attach to purpose, not just role.
Scope every access
Clinicians see what care needs; billing sees codes not notes; research sees tokenized extracts. Minimum necessary, enforced.
Export the evidence
Access reviews and auditor requests resolve to per-record trails: who, what, when, and under which purpose.
Asked by teams like yours.
Do you sign a BAA?
Yes — business associate terms are available under agreement, scoped to the vault’s PHI processing.
How does research de-identification work?
Extracts replace direct identifiers with tokens and suppress or generalize quasi-identifiers per your protocol, before data leaves the vault.
What about patient access and erasure requests?
Lookup plus token-lifecycle deletion gives you a repeatable fulfillment workflow instead of a ticket queue.
See it on your own workflow.
Bring one sensitive flow. We will map it to a vault schema live in 30 minutes.
One record, three legitimate views.
A single request carries its purpose; care, billing, and research each receive only what their purpose allows — and every view is audited.
treating team"] P -- billing --> B["Codes, not notes
masked fields"] P -- research --> RE["Tokenized extract
+ redaction"] C --> A["Audit event"] B --> A RE --> A