Partitioned per org and workspace
Every org's and workspace's data sits in its own partitions, and the server checks the scope on every request. Enterprise gets a dedicated stack.
What is in place today. Each item is something the platform does, not a promise on a roadmap.
Every org's and workspace's data sits in its own partitions, and the server checks the scope on every request. Enterprise gets a dedicated stack.
IAM-style Allow and Deny policies, checked on the server for every call, from the app, an agent or MCP.
Console SSO with Okta or Google, and workspace sign-in with Okta, Google or a password, with domain verification.
AES-256 field-level encryption for sensitive fields, encrypted workspace secrets, TLS in transit.
Every table has point-in-time recovery turned on, so data can be restored to any second in the recovery window.
AI requests are redacted and stored encrypted for 30 days, then deleted. The billing ledger keeps numbers and hashes, never text.
Delivery traces are redacted before they are stored and expire automatically.
Glyphs are hash-checked before every run, limited to allowed domains, and keep secrets out of logs.
MCP calls are checked against the caller's policy and audited. Only a person publishes to production.
Health heartbeats from every stack feed a public status page.
Connect your AI client, build your backend, ship to production.