SSO & SCIM, on your IdP
Single sign-on over SAML 2.0 or OIDC, plus SCIM provisioning, ride your existing identity provider. Add or remove someone there and their Beemy seat follows.
Enterprise & Platform
SSO, admin controls, compliance readiness and region residency for teams of 50+ — and a platform to build on, all resting on the isolation Beemy already enforces for one person. Nothing here is softened to get you in the door.
For teams of 50+
SSO, admin controls, custom integrations, compliance reviews and white-glove onboarding for teams of 50+. Custom pricing, human conversation — the details below are what that conversation covers.
Identity & administration
Single sign-on over SAML 2.0 or OIDC, plus SCIM provisioning, ride your existing identity provider. Add or remove someone there and their Beemy seat follows.
Offboard someone in your IdP and their access is revoked on the next check — no long-lived grant sitting around to claw back later.
The admin boundary
An admin manages membership, seats and policy, and can read audit metadata for the team silo — who did what, when, at a category level. An admin is never a path into a member's personal-silo content. That isn't a policy we ask you to trust; it's not a capability the architecture hands them in the first place. Same wall, drawn around your whole team.
Compliance
Most of SOC 2 Type II, for us, is naming and evidencing controls the architecture already enforces — not building new ones from scratch. Type II is a window, not a snapshot: it's earned by an auditor observing those controls hold over time, and that observation window hasn't opened yet. Plainly: Beemy is readiness-stage, not yet certified.
Personal, work and public knowledge live in physically separate stores — the same wall described on our Trust page, not a setting that could be misconfigured.
Data is encrypted with per-tenant keys wrapped by a managed key hierarchy, not one shared secret protecting everyone at once.
Every sensitive action is appended to a hash-chained log — entries can't be edited or deleted after the fact, only appended.
Deleting an account destroys the encryption keys that protect it, not just a database row — the data becomes unrecoverable by construction.
Inference requests are routed so your content is never used to train shared models — the same commitment as our Trust page, enforced at the routing layer.
Every byte moving between you and Beemy, and between Beemy's own services, travels over TLS 1.3.
Every change ships through automated security checks before it reaches production — not a periodic review, a gate that runs on every commit.
The controls above run today. What's ahead of us is the audit itself — engaging an auditor and completing the observation window Type II requires. See the full commitments on our Trust & Security page, or read the honest version of "enterprise-ready".
Data residency
Beemy is built to pin a region-pinned tenant's data, encryption keys and model inference in-region — the same silo wall and row-level isolation that separates personal from work, drawn around a region instead of just around you. It's an enterprise-deployment capability we scope in the conversation below, not a default switched on for every account today.
Talk to us about your region →For builders
Beemy is built as a platform, not just a product. A credential is scoped to a single silo — it can mint a context for that silo alone — and every call runs through the same gateway and policy engine the app itself uses. A write still waits for your approval until you raise the dial on that surface, and a credential can never grant itself more autonomy than you've handed it.
A skill is a when-this-then-that binding of triggers Beemy already watches to actions it already takes — no new execution primitive, just your own automation on top of what exists. A skill lives inside exactly one silo, can't reach across the wall, and only acts at the autonomy level you've already granted that surface. Authoring one never grants it more power than you have.
Questions
Not yet — we're readiness-stage. The controls SOC 2 Type II evidences (silo isolation, encryption, audit logging, deletion) are already built and enforced in the architecture. What's left is the audit itself: engaging an auditor and completing the observation window Type II requires, which hasn't opened yet.
No. An admin manages your team's seats and policy, and can see audit metadata for the team silo — who did what, when, at a category level. An admin has no path into a member's personal silo. That's true by construction, not by a promise we're asking you to trust.
For enterprise tenants, we can pin a tenant's data, encryption keys and model inference to a region — the same silo wall and isolation, drawn around a region instead of just around you. It's a deployment we scope in the enterprise conversation, not a default flipped on for every account.
There's no public API live yet. What's specified is the model it will ship on: a credential scoped to a single silo, routed through the same gateway and policy engine the app itself uses, and unable to grant itself more autonomy than you've handed it.
A skill is a user-authored automation — when this happens, do that — built from triggers and actions Beemy already has, not a new capability. A skill lives inside exactly one silo, can't reach across the wall, and can only act at the autonomy level you've already granted that surface.
Talk to us. Most enterprise requirements get worked out in that conversation directly, rather than in a self-serve form — that's true for custom integrations, compliance reviews and white-glove onboarding alike.
Beemy is in private beta. Join the waitlist and go from connect to a quiet, triaged inbox — in your first session, not a week later.