Agent Governance for Healthcare Engineering Teams
Health software teams adopt coding agents at the same pace as everyone else, under a constraint the rest of the market does not carry. The systems those agents touch hold patient information, and the tools that can reach those systems are the tools already through security review. Modulo is built to be that kind of tool: self-hosted, with the record and the data on your infrastructure.
Data placement
Modulo runs on your infrastructure. With default settings and no connectors configured, it makes no external network calls. Prompts, outputs, run records, and audit events are written to your Postgres database, and encrypted credentials are keyed by FERNET_KEY in your environment rather than a vendor's.
The model decision stays with you
Where a workflow touches protected health information, the model provider is a decision your team makes and an agreement your team holds. Modulo routes to the backend you configure: a hosted provider under your contract, or a local OpenAI-compatible server when inference needs to stay on your network.
A record for a change to clinical or patient-facing software
Every agent action appends an immutable audit event, chained with the SHA-256 hash of the previous event, append-only at the application and database layers. Each run records the prompt, the output, the schema check, and the approval decisions, so a change to a patient-facing service has a record of who approved it and what the agent did.
Access and approvals
Single sign-on over OIDC or SAML, just-in-time provisioning, identity-provider groups mapped to teams, and four organization roles. Human approval gates can pause any step, and a human-only gate cannot be claimed or decided by an agent. Single sign-on, team roles, and the audit viewer and export are part of the paid Team tier; the Community tier runs the pipelines and records the audit chain without them.
Fit to HIPAA
HIPAA obligations land on covered entities and their business associates. The controls an auditor and a security officer look for are access control, an audit record, data location, and a defined retention policy. Modulo is the mechanism for those controls in the deployment you assemble; the assessment, the policy, and the agreements are yours.
Placement is self-hosted, with your database and your keys. How the split works.
Related use cases
Audit Trail for AI-Generated Code
A hash-chained, append-only audit trail plus a versioned snapshot for every run, so you can prove who approved what an AI agent changed.
Own Your LLM Spend: Bring Your Own Model Keys
How Modulo routes agent calls through your own provider credentials, accounts for tokens per run, and leaves the model contract with you.
Self-Hosted Agent Governance
What self-hosted agent governance is, why regulated teams keep the control plane on their own infrastructure, and how Modulo is deployed.