The administrative backbone
your apps build on

People, events, documents, billing, and communications — modeled your way with tenant-defined schemas, automated with workflows and AI, and connected to email, calendars, and meetings, behind one secure multi-tenant API.

Schema on your terms

Every entity type is composed from tenant-defined, versioned attributes. Model your domain without waiting on a platform release.

Documents done right

Immutable versioning, check-out/check-in, role-based access, a full audit trail, and built-in full-text search across everything you store.

Events & scheduling

Meetings, deadlines, and to-dos with participants, roles, and RSVP status — queryable by time range out of the box.

Billing built in

Invoices with lifecycle control, per-tenant numbering, line items generated straight from events, and payment records that keep the books straight.

Communications

Internal messages and system notifications with threads, read state, and a unified inbox — triggered automatically by platform activity.

Workflows

“When this happens, run these steps” — configured per tenant, fired by platform events, schedules, manual runs, or webhooks, with retries built in.

AI tasks

Summarize, extract, classify, and generate over your own documents, with schema-validated output and per-tenant metering. Opt-in, on the model account you choose.

Email, in and out

Communications become real email for people who opt in; inbound mail becomes documents with matched participants — through the vendor you bind, never a hardwired one.

Calendar & meetings

Invite the platform to a meeting: the entry mirrors as an event, a notetaker attends, and the transcript lands as a document that your workflows turn into notes, hours, and follow-ups.

Multi-tenant to the core

Hard tenant isolation on every record and every request. Registered apps authenticate with scoped API keys; end users sign in with bearer tokens bound to one person.

Learn it, then look it up

Start with the mental model, make your first call in the quickstart, then go task by task with the how-to guides. The full OpenAPI reference has an interactive explorer — and every docs page is plain Markdown underneath (llms.txt), so AI agents can read it all in one pass.

Open the API reference →