For enterprise

Semantic search across SharePoint, Drive, Jira and the cloud — without sending it all to a SaaS vendor.

Read-only credentials, isolated databases per source, encryption on the roadmap, no telemetry. StrataFS is the search layer your compliance team will actually approve.

The problem

Enterprise knowledge lives in SharePoint sites, OneDrive, Google Drive, Jira projects, S3 archives, and engineering laptops. Indexing all of it with a SaaS vendor means sending it all to a SaaS vendor — a long conversation with security, a longer conversation with legal, and a monthly bill that scales with corpus size.

StrataFS is the inverse. One open-source binary, deployed inside your perimeter, that speaks to all those sources with read-only credentials and keeps the index on your own disk.

What it indexes

Compliance posture

Air-gapped deployments: StrataFS runs without any outbound connection once the embedding model is downloaded. Bundle the Docker image + the ONNX file, deploy inside the network, point at internal S3-compatible storage.

Deployment shapes

Source-level access control

Today, sources are configured in config.toml — anyone with access to the binary can query any configured source. RBAC with source-level permissions is in active development and is the next major release item.

Procurement

StrataFS is MIT-licensed. No procurement to do — clone the repo, run the binary, you're in production. For organizations that need a commercial support agreement, contact stratafs@neullabs.com.

Bring all your enterprise knowledge under one search.

Read-only. Self-hosted. Compliance-friendly. MIT licensed.