Documentation
How it works,
in detail.
Hyves is in early access, so the in-product help centre is still being written. What exists today is arguably better: long-form documentation of how the system actually runs, written to be argued with rather than skimmed.
Available now
Start here.
The method, end to end
The complete operating procedure: ten steps across map, enrich, engage, and close — with the signal taxonomy, the verification waterfall, and where every human approval gate sits. The most detailed account of how Hyves runs.
Read the methodPlatform architecture
How one shared account memory sits under every module, why that changes what the AI can do, and what each surface is responsible for.
Platform overviewSYNC and the agent swarm
How missions are briefed in plain language, decomposed into parallel worker runs, and held in the approval queue before anything leaves the building.
About SYNCIntegrations
What connects today — CRM sync, email and calendar, data and enrichment partners — and how the account record stays the source of truth across them.
Browse integrationsData handling and security
Lawful basis, tenant isolation at the database layer, EU hosting, deletion on request, and which sources the engine is allowed to read.
Trust & securityBeing written
What lands with general availability.
- Step-by-step setup guides with screenshots
- Full endpoint reference with request and response examples
- TypeScript and Python SDKs
- Recipes: common missions, campaign patterns, and approval workflows
We'd rather say that plainly than publish a hundred stub articles. Early-access teams get the same answers directly from the people who built the system.
A question the docs don't answer yet?
Ask SYNC — the guide in the corner of this page knows the platform and can take you straight to the right section. For anything deeper, a human answers.