Skip to main content

Workflows

Workflows are the core of Splice automation. They define the sequence of actions that connect your logistics systems, automate data processing, and trigger notifications based on supply chain events.

What is a Workflow?​

A workflow is a series of connected nodes that:

  • Trigger on events (new shipments, schedule, incoming emails)
  • Process data (transform, filter, validate information)
  • Take actions (send notifications, update systems, generate reports)
  • Connect systems (sync data between TMS, WMS, carrier APIs)

Creating Workflows​

Visual Workflow Builder​

Splice provides a drag-and-drop interface to build workflows without coding:

  1. Start with a trigger - Choose what initiates your workflow
  2. Add processing nodes - Transform and route your data
  3. Connect action nodes - Define what happens with the processed data
  4. Test your workflow - Validate logic before going live
  5. Deploy and monitor - Activate your automation

Common Workflow Patterns​

Shipment Tracking Automation

Cron Trigger → Carrier API → Process Updates → Notify Customers → Update TMS

Exception Handling

Carrier Webhook → Check Delays → If Delayed → Send Alert → Create Task

Data Synchronization

TMS Export → Transform Data → Validate → Update WMS → Report Results

Workflow Components​

Triggers​

Workflows start with triggers that detect events:

  • Cron/Schedule - Run at specific times or intervals
  • Webhooks - Respond to external system events
  • Email - Process incoming emails with attachments
  • File uploads - React to new files in connected systems
  • API calls - Start workflows via Splice API

Processing Nodes​

Transform and route data through your workflow:

  • Code nodes - Custom JavaScript for complex logic
  • If/Switch nodes - Conditional routing based on data
  • Transform nodes - Map and modify data structures
  • Filter nodes - Include/exclude data based on criteria
  • Merge nodes - Combine data from multiple sources
  • Loop nodes - Cycle through data sets one item at a time

Action Nodes​

Perform operations with processed data:

  • HTTP requests - Call external APIs
  • Email notifications - Send alerts and reports
  • File operations - Create, read, or upload files
  • Database operations - Store or retrieve data
  • Logistics connectors - Use pre-built integrations

Leverage AI​

Define an agent as step in a workflow task:

  • Use AI agents for natural language processing
  • Extract data from unstructured documents
  • Automate decision making based on historical patterns
  • Generate responses and content dynamically
  • Classify and route items using machine learning
  • Enhance data quality through AI validation

Executing Workflows​

Automatic Execution​

Workflows run automatically when triggered:

  • Real-time triggers respond immediately to events
  • Scheduled triggers run at predetermined times
  • Webhook triggers activate when external systems send data

Manual Execution​

You can also run workflows manually:

  • Test runs during development
  • One-time operations for data migration or cleanup
  • Debug executions to troubleshoot issues

Execution Context​

Each workflow execution includes:

  • Unique execution ID for tracking
  • Input data that triggered the workflow
  • Execution timestamp and duration
  • Node-by-node results showing data flow
  • Success/failure status for each step

Monitoring Workflow Executions​

Execution Dashboard​

Track all workflow activity in real-time:

  • Recent executions with status indicators
  • Success/failure rates over time
  • Performance metrics (execution time, throughput)
  • Error summaries for failed executions

Execution Details​

Drill down into individual workflow runs:

  • Step-by-step execution log showing data at each node
  • Error messages with stack traces for debugging
  • Input/output data for each processing step
  • Execution timeline showing bottlenecks

Alerts and Notifications​

Stay informed about workflow health:

  • Success notifications for critical operations
  • Failure alerts when workflows encounter errors
  • Custom alerts based on workflow output data

Best Practices​

Design Principles​

  • Keep workflows focused - One workflow per business process
  • Handle errors gracefully - Add error handling and recovery logic
  • Use descriptive names - Make workflows self-documenting
  • Test thoroughly - Validate with real data before deployment

Performance Optimization​

  • Minimize API calls - Batch requests when possible
  • Use filters early - Reduce data processing overhead
  • Monitor execution times - Optimize slow-running workflows

Maintenance​

  • Regular testing - Ensure workflows still work as expected
  • Update credentials - Keep API access current
  • Review logs - Check for patterns in failures or performance
  • Document changes - Track modifications and their purposes

Troubleshooting​

Common Issues​

  • Authentication failures - Check credential validity
  • Timeout errors - Increase timeout settings or optimize logic
  • Data format mismatches - Verify input/output data structures

Debugging Tools​

  • Execution logs - Detailed step-by-step information
  • Test mode - Run workflows with sample data
  • Node inspection - View data at any point in the workflow
  • Error notifications - Immediate alerts when issues occur