Business Flow Quick Start

Business Flow is a workflow application designed for batch processing scenarios. It is suitable for building one-time automated tasks such as document processing, content generation, and data transformation.

This tutorial will guide you from scratch to build a Document Content Processor: receive a document uploaded by the user, automatically extract the text content, process it through a large language model, and output the result.

The diagram below shows the complete flow for reference at any point during the build:

Step 1: Create a Business Flow

1. Go to AIGC Apps / App Center and click the "Create App" button
2. Select Business Flow as the application type, fill in the application name (e.g., My Business Flow), and click "Save & Create"
3. The system automatically redirects to the workflow canvas where you can start building

Step 2: Configure the Start Node

⚠️ Keep the default values.

🚨 Give nodes and variables clear, descriptive names for easier reference and troubleshooting later.

Click the Start Node on the canvas and add an input variable in the right configuration panel:

Variable Name	Type	Description
input	File	The document file uploaded by the user

Step 3: Add a Text Extraction Node

Add a Text Extraction node after the Start node to parse the document content into plain text:

1. Click + after the Start node and select the "Text Extraction" node
2. Input Variable: Select the input variable configured in the Start node
3. Document Parsing Service: Select an available parsing service from the dropdown
4. Image Parsing: Decide whether to enable this as needed (when enabled, images within the document can also be parsed)

Step 4: Add a Model Node

Add a Model Node after the Text Extraction node to process the extracted text with AI:

1. Click + after the Text Extraction node and select the "Model" node
2. Select Model: Choose an appropriate LLM from the dropdown
3. Input Variable: Select the output variable from the Text Extraction node
4. Prompt: Fill in the system prompt based on your business needs. Example:

You are a professional document analysis assistant. Please analyze and summarize the following document content with these requirements:
1. Extract the key points with a clear structure
2. Use conc and avoid redundancy
3. Highlight any data or conclusions

Document content: {{input}}

Step 5: Add an End Node

Add an End Node after the Model node to configure the final output:

1. Click + after the Model node and select the "End" node
2. Output Variable: Select the output variable from the Model node as the final result

Step 6: Debug and Publish

Debug

Click the "Debug" button in the upper right corner of the canvas. In the right debug panel, upload a test file and observe whether the flow executes as expected:

• After each node executes, input/output details are displayed to help troubleshoot issues
• If the text extraction result is empty, check whether the document format is supported
• If the model response is inaccurate, adjust the prompt content or switch to a different model

Publish

After debugging passes, click the "Publish" button in the upper right corner and select a publish option (Publish Update / Run)

Complete Flow Diagram

Start Node (input: File)
    ↓
Text Extraction Node (parse document content → text)
    ↓
Model Node (process text content → result)
    ↓
End Node (output result)

Run History

Each execution of a Business Flow generates a run record. Click the Run History icon in the upper right corner of the application to view all historical execution records, including:

• The input file and output result for each execution
• Execution details and duration for each node
• Execution status (success / failure) and error messages