Where Samskara runs, and where your data doesn't.
SAMSKARA is a control plane, not a data store. Here's what that means for your data, your team, and your security posture.
Where does SAMSKARA run?
Entirely on infrastructure you control — a VM, on-prem hardware, or your own cloud VPC. A single docker compose up brings up the full stack as containers you own. Nothing runs in a SAMSKARA-operated cloud.
Where does my data live?
Notebook workspace files live on disk under your control, scoped per organization and project. ArrowLake tables land in your own object storage as open Iceberg data you can read with any compatible tool. Data you query through a connected database stays in that database — SAMSKARA never copies it into a separate managed store.
Does data ever leave my network?
Only for AI Sparkle, and only if you choose to point it at an external model provider. Route it at a local model instead — Ollama works out of the box — and every prompt and completion stays inside your own network.
Where are secrets stored?
Org secrets are encrypted at rest using a key you generate and control, stored in your own database. Values are decrypted and injected into the runtime environment only at execution time — the API never returns a secret's raw value.
How does AI connectivity work?
AI Sparkle assembles context — live session variables, ArrowLake schemas, the sm.* API surface, and examples from your training history — server-side, then forwards a single request to whichever AI endpoint you've configured. The model itself never has standing access to your databases or storage.
How are notebook sessions isolated?
Each session runs in a fully isolated environment — packages you install in one notebook can't affect another, and session state never leaks across users or projects. Enable the Docker runtime profile and each session gets its own container-level boundary.