When your permissions and relationships outgrow what tables can hold
Permissions, access and relationships modeled as a graph — where a relational database would need a web of join tables nobody understands. Regulated, with a clear access boundary.
How we do it
Relationships and permissions as a graph: “who may see what” is a path through nodes, not a join across five tables — so a new rule is a new edge, not a new migration.
Access resolved by type and role where a graph is not needed — 500+ routes with permissions that do not scatter across the code.
Regulated domains (HIPAA, finance): a record that survives review and a clear boundary of who may do what.
Proof
Permissions as graph edges, realtime transcription with speaker separation — a HIPAA platform, product sold.
Permissions as a graph on a HIPAA platform · Read the case study →How to engage
A fixed-scope project. NDA as standard. We reply within one business day.
- Neo4j
- WebRTC
- AWS Transcribe
- Node.js
- PostgreSQL
- AWS