Everyone
Understand FlowOS before you automate anything
FlowOS Automation Hub turns repeatable business work into visible, governed workflows. An event starts the work, rules choose the path, actions create the outcome, and history explains what happened. FlowOS is deliberately broader than a simple “if this, then that” tool: it also includes approvals, lead capture, project planning, connector evidence, tenant administration, and operational health.
Before you start
- No technical knowledge is required.
- Have a business process in mind that currently involves repeated handoffs.
#Choose the path that matches your job
Everyday user
Start with notifications, approvals, tasks, and execution results. You do not need to understand databases or webhook signatures.
Workflow builder
Learn triggers, conditions, actions, waits, testing, publishing, and replay. Begin with one observable outcome.
Tenant administrator
Configure people, roles, settings, integrations, security, backups, and readiness for one workspace.
Platform administrator
Manage workspaces, plans, gateways, system health, and production infrastructure across tenants.
Developer
Use scoped API keys, OpenAPI, signed webhooks, connector operations, queues, and idempotency.
Buyer or reviewer
Follow the guided story to see how the modules form one accountable business journey.
#Know the product boundaries
FlowOS coordinates work; it does not replace every specialist CRM, calendar, finance, or project application. Use native connectors or typed REST operations to connect those systems. A feature marked Configuration required needs provider or environment setup. A feature marked Not production-certified must remain outside production claims until its release evidence is complete.
#Aim for one safe first success
- Describe one event in plain language, such as “a qualified lead submits the website form.”
- Choose one visible outcome, such as a review request or assigned task.
- Build the smallest path that connects the two.
- Test with fictional data and inspect the execution record.
- Publish only after permissions, workers, and failure ownership are clear.
#Plain-English glossary
- Action
- A workflow step that creates an observable outcome.
- Approval
- A recorded human decision before guarded work continues.
- Central platform
- The workspace registry and platform-level commercial and operational controls.
- Connector
- A tenant-scoped, encrypted connection to another system.
- Execution
- One recorded run of a workflow definition.
- Idempotency
- A guarantee that repeating the same request does not repeat the business effect.
- Readiness label
- A factual status describing whether a capability is available, needs configuration, or lacks production certification.
- Replay
- A new auditable execution based on an earlier run for comparison or correction.
- Scheduler
- The once-per-minute process that starts due recurring and delayed work.
- Tenant workspace
- The isolated users, settings, workflows, records, and integrations belonging to one organization.
- Trigger
- The event that starts a workflow.
- Worker
- A supervised background process that executes queued work.
#Common mistakes
- Trying to automate an entire department in the first workflow.
- Treating an installed connector as proven production support.
- Publishing before a queue worker and scheduler are supervised.
#Verify the result
- I know which role-based path applies to me.
- I can explain the first event and expected outcome.
- I understand that readiness labels are evidence-based.