Projects
Structured containers for work. Every project is rooted in a workspace, anchored to a blueprint, and owned by a person.
Projects with operational backbone
Not folders. Projects in Rapidesk carry ownership, blueprint lineage, policies and a fully addressable URL.
- Bi-directional structure — every task knows its project, workspace and blueprint
- Public or workspace-scoped — share without losing tenant boundaries
- Snapshot history — restore any project to a previous operational state
Tasks
Tasks aren't checkboxes. They're operational atoms — owned, scoped, addressable, and reasoned over by every other surface.
Tasks with provenance
Every task records why it exists, what blueprint it came from, and what policies apply. The whole lifecycle is one timeline.
- Owners, reviewers, watchers — all first-class roles
- SLAs, dependencies and cycle-time targets per task
- Subtasks roll up to parents automatically
Workspaces
Tenant-aware surfaces with their own data, permissions, branding and AI context. Built for organisations that need real boundaries.
True multi-tenancy, not folders
Workspaces are isolated by design. Data, search, notifications and AI all respect tenant boundaries without any extra config.
- Per-workspace identity providers (SCIM, SAML, OIDC)
- Workspace-scoped Vault, blueprints and policies
- Guest workspaces for partners and vendors
Blueprints
Operational templates that encode how your business actually works. Spin up new projects from your own institutional knowledge.
Blueprints, not templates
Blueprints stay alive after instantiation. Update one and every downstream project picks up the change with versioned diffs.
- Versioned, auditable, owned by a team
- Auto-propagation with operator-controlled diffs
- Composable — one blueprint can include another
Permissions
Permissions are operational. Roles map to real business rules, not generic ACLs.
Role-based by design
Roles are first-class. Custom roles compose permissions across projects, blueprints, the Vault and admin surfaces.
- Granular scope: workspace, project, blueprint or surface
- Inheritance with explicit overrides
- Just-in-time access requests with audit
Notifications
Less noise, more signal. Notifications respect your operational context and your right to focus.
Focus-first inbox
Notifications collapse by context. Anything not actionable becomes a digest. Quiet hours, smart routing and digest scheduling are first-class.
- Per-surface preferences with smart defaults
- Auto-grouping by project, blueprint or person
- Quiet hours synced with your calendar
AI
Agents that reason over your operational model — projects, blueprints, permissions, history.
Operational AI agents
Embedded in every surface. Inherits the surface's permissions, data scope and operational context.
- Suggested briefs, drafts and routing actions
- Workspace-scoped — your data never leaves its tenant
- Tunable: choose which surfaces and roles get AI
Automations
Triggers, conditions and actions over the operational graph. No code required, code allowed when you want it.
Trigger studio
A real automation surface — typed inputs, branching, and a visual graph. Run on the operational model, not on document events.
- 200+ triggers across projects, blueprints, the Vault
- Conditional branches with typed comparisons
- Code blocks — JavaScript with sandbox
Operational visibility
Dashboards, audit, alerts and analytics — built into the model, not bolted on as a separate product.
Visibility you can act on
Every dashboard ties back to the same operational graph. Click a metric to land in the work it represents.
- Cross-workspace rollups for ops leads
- SLA & cycle-time alerts with playbooks attached
- Audit log retention up to 7 years