AI GOUser Guide

Workflow Automation Basics

When enterprise colleagues perform daily operations on their respective Custom Apps (such as accounting staff registering a new voucher, or sales staff submitting a new order), these data are written to the backend Data Center. The "Workflow Builder" of AI GO allows administrators and developers to establish background automation rules to automatically execute cross-system notifications or database table updates when these data change, eliminating manual repetition.

Positioning of Automated Workflows

Workflow automation is a monitoring and execution mechanism running in the system background. Its core logical rule is: "When a specific table in the Data Center changes (Trigger Condition), automatically execute subsequent tasks (Execution Action)."

Because daily operational colleagues work in independent Custom Apps, automated workflows act as "invisible glue" to connect data generated by different Apps, achieving seamless automation of enterprise operations.

Configuring Trigger Conditions and Execution Actions

A complete automated workflow consists of at least one "Trigger Condition" and one or more "Execution Actions":

1. Trigger Condition (Trigger)

The trigger condition is the starting point of the automated workflow. AI GO provides three main trigger types:

  • Database Event Trigger: Activated when specific tables in the Data Center change. For example, "When a new record is created in the Sales Order Table (a colleague submitted an order via the Sales App)" or "When the status of the Accounting Table is updated to 'Paid'."
  • Time Trigger (Scheduled Task): Starts at specific points in time or periodically. For example, "Extract data from the daily sales table automatically every night at 9 PM."
  • Manual Trigger: Actively triggered by users clicking buttons in specific Custom App interfaces.

2. Execution Action (Action)

Steps automatically executed after trigger conditions are met. Common actions include:

  • AI Processing Action: Send changed data to AI for analysis. For example, "When a new lead table record is created, let AI automatically summarize its core needs and determine client intent."
  • Internal Data Center Action: Add or modify data in other tables within the Data Center. For example, "When delivery is confirmed, automatically reduce the quantity in the Stock & Warehousing table."
  • External Communication Notification: Send notifications to external platforms. For example, "Send a LINE message to the responsible salesperson" or "Send an email to the purchasing manager."

Steps to Create an Automated Workflow

Taking "When a colleague changes product stock in a Custom App and the stock falls below the safety threshold, automatically send a LINE notification to purchasing staff" as an example:

Step 1: Create a New Workflow

  1. Click "App Integration" or "Workflow Automation" under the "AI Applications" section in the left navigation bar of the Dashboard.
  2. Click "Create Workflow" and enter a name such as "Inventory Shortage Auto-Alert Workflow".

Step 2: Set the Trigger Source

  1. Click the trigger node and select "Database Event Trigger".
  2. Select the underlying "Stock Table" from the Data Center.
  3. Select "On Update" as the event type.
  4. Set the filter condition: "When 'Current Stock' is less than 'Safety Stock'."

Step 3: Add AI Analysis Step (Optional)

To make notification content more valuable, we can add an AI processing node:

  1. Add a node and select "AI Processing".
  2. Set the prompt: "Based on the triggered inventory change data, organize the product name and current stock, and briefly analyze the average monthly sales of the past three months so that purchasing colleagues know the urgency of restocking."

Step 4: Configure Notification Action

  1. Add a node below the AI node and select "Send External Notification".
  2. Select "LINE" or "Email" as the communication software.
  3. Set the recipient as the responsible purchasing department group.
  4. Insert the AI analysis text generated in the previous step into the message body.

Step 5: Test and Enable

  1. Click the "Test Run" button in the upper right corner to simulate database changes and ensure LINE receives the notification correctly.
  2. Once verified, toggle the switch at the top of the workflow to "Enabled".
  • Cross-App Sales Alert: When sales colleagues add a large order in the "Sales App", automatically update the "CRM Table" in the Data Center and pop up a reminder on the manager's messaging software.
  • Daily Sales Report: Scheduled to trigger every night at 9 PM, AI automatically extracts data from the Data Center's "Sales Order Table", analyzes totals, and sends a daily sales summary to the management inbox.
  • Inquiry Auto-Distribution & Notification: When external clients submit needs via the "Online Lead App", data is written to the "CRM Table" of the Data Center. This triggers the workflow where AI analyzes client needs and distributes them to the most suitable salesperson.