- Introduction
- Getting started
- Build on any canvas
- Build with Case
- Reference
- Defining case keys (system vs. external)
- Establishing task I/O and write-back contracts
- Modeling primary and secondary stages
- Triggering a case from Data Fabric
- Implementing stage-level personas and permissions
- Setting SLAs and automated escalation rules
- Configuring a rework loop (re-entry)
- Configuring and testing the Case Manager Agent (preview)
- Build with Flow
- Build with BPMN
- Operating
- Monitoring
- Optimizing
- Reference information
Flow use cases
Six processes built on the Flow canvas, from order intake to screening checks, with the nodes each one uses.
These processes are built on the Flow canvas. In each one, something comes in, an AI agent decides what to do, and APIs carry out the decision. They are examples, not step-by-step instructions.
Order intake and enrichment
What happens: an order or a sales lead arrives. An agent validates it, checks stock and the customer's credit over API, and creates the order in the enterprise resource planning (ERP) or customer relationship management (CRM) app.
You use: a connector trigger, HTTP Request, an Autonomous agent, a Decision, and connectors.
Claim triage
What happens: a first notice of loss arrives, which is the first report of an insurance claim. An agent classifies its severity, checks the coverage against the policy system's API, and routes the claim to the right adjuster's queue.
You use: a connector trigger, an Autonomous agent, HTTP Request, a Switch to route by severity, and Queue tools.
Service desk request fulfillment
What happens: an employee asks for help, for example a password reset or access to an app. An agent reads the request, pulls the employee's context from the identity, IT service management, and human resources (HR) systems over API, decides what to do, and carries it out.
You use: an Autonomous agent or a Conversational agent (Preview), HTTP Request, and connectors. Build a chat agent workflow shows the chat version step by step.
Invoice intake
What happens: an invoice arrives as a document. An extraction step reads its fields, an agent checks them against the purchase order over API, and the invoice is posted when it matches.
You use: Extract (Preview), an Autonomous agent, HTTP Request, and a Decision.
Transaction reconciliation
What happens: APIs pull both sides of a transaction, for example the bank record and the ledger entry. When they don't match, an agent explains the difference and proposes a fix, and the process either applies the fix over API or hands it to a person.
You use: HTTP Request, an Autonomous agent, a Decision, and connectors.
Screening check
What happens: a new customer signs up. An agent runs sanctions and adverse-media lookups over API, scores the result, and writes the score back to the customer record. Customers with a match are flagged for review.
You use: a connector trigger, HTTP Request, an Autonomous agent, and a Decision.
Related pages
- Your first Flow: build a working process on the Flow canvas.
- Node reference: every node these use cases mention.
- Choosing a canvas: build the same processes on the Case or BPMN canvas.