Yeeflow AI Builder helps you create application resources using natural language. Instead of configuring everything from scratch, you can describe your business requirements and let AI propose an implementation.
You stay in control throughout the process: review the plan before creation, check the generated resource, and continue refining it with AI.
Supported Resource Types
You can use AI-assisted creation for the following 11 resource types:
Resource type | What to describe in your request |
Approval form | Information to collect, approval steps, and business rules |
Data list | Data to store, required fields, and how the information will be used |
Document library | Documents to manage and the information needed to organize them |
Form report | Source approval form and the submitted information you want to report on |
Data report | Source data and reporting requirements |
Dashboard | Target audience, key metrics, charts, and data sources |
Scheduled workflow | Run schedule, conditions, data sources, and actions |
Knowledge source | Information sources and the knowledge you want to make available |
AI Agent | Responsibilities, expected inputs and outputs, and tasks to perform |
Copilot | Intended users, questions to answer, and tasks to support |
Custom Service | Reusable business logic, expected inputs and outputs, and integration requirements |
These are existing Yeeflow resource types that support AI-assisted creation.
Before You Start
Open the application where you want to create the resource. Make sure you have permission to add resources to that application.
If your resource depends on existing data or configuration, identify the relevant sources before starting. For example, a report may require an existing approval form or Data List.
Step 1: Open New Application Resource
In your application, open New application resource.
In the AI creation area, select the resource type you want to create and enter your requirements.
Step 2: Describe What You Want to Build
Explain the business purpose and the details the resource needs.
For example, to create a supplier Data List:
Create a supplier management Data List. Include supplier name, contact person, contact email, supplier category, contract amount, cooperation status, and contract renewal date. Supplier category should include Materials, Equipment, and Services. Cooperation status should include Pending, Active, and Inactive.
For better results, include:
Purpose: What business problem should the resource solve?
Required information: Which fields, metrics, inputs, or outputs are needed?
Business rules: What conditions or processing rules should apply?
Existing resources: Which lists, forms, or other resources should it use?
Expected result: What should users be able to view or do?
Click Start with AI to submit your request.
Step 3: Review the Implementation Plan
AI analyzes your request and prepares an Implementation Plan before creating the resource.
Review the plan to check whether it matches your requirements. The details depend on the selected resource type and may cover the proposed structure, configuration, layout, or processing logic.
Pay particular attention to:
Review area | What to check |
Business purpose | Does the proposal address the task you described? |
Structure | Are the required fields, components, or steps included? |
Data sources | Does the plan refer to the correct existing resources? |
Business logic | Are the conditions, actions, and processing rules appropriate? |
Missing details | Is anything required for your scenario still unclear or absent? |
If changes are needed, explain them in the conversation before confirming the plan.
For example:
Make supplier name required. Add a contract renewal date field, and use Active, Pending, and Inactive as the cooperation status options.
Review the revised plan before proceeding.
The resource is created only after you confirm the Implementation Plan.
Step 4: Confirm and Create the Resource
When the plan meets your requirements, confirm it to proceed with creation.
AI then generates the application resource based on the confirmed plan. Follow any additional setup prompts shown for the selected resource type.
Step 5: Check the Generated Resource
After creation, open the resource and review its configuration.
Depending on the resource type, you may need to:
Check field names, types, and required settings.
Review the form or dashboard layout.
Verify report sources and displayed information.
Check workflow schedules, conditions, and actions.
Review Agent or Copilot instructions and connected resources.
Test Custom Service inputs, outputs, and logic.
Complete any remaining configuration and test the resource before making it available for business use. Follow the saving, publishing, or enabling process for that resource type.
Continue Editing with Ask AI
After the resource is created, use Ask AI to return to the original saved conversation and continue refining your requirements.
For example:
Add a contract renewal date field to the supplier list.
Or:
Update the dashboard to include a monthly comparison of contract amounts.
Returning to the saved conversation lets you continue from the earlier discussion without describing the entire requirement again. Review the proposed changes before accepting them.
Refine Designs with Build with AI
For supported resources, you can also make targeted changes directly in the designer using Build with AI.
This is available for:
Approval forms
Data List forms
Public forms
Dashboards
Scheduled workflows
To make a targeted change:
Open the relevant form or workflow designer.
Switch to Build with AI.
Select the control or workflow node you want to change.
Choose Add to chat to include it as context.
Describe the required modification.
Review and confirm the AI changes.
For example, after adding a form control to the conversation:
Move this field into the Supplier Details section and make it required.
Or, after adding a workflow node:
Update this step to send the reminder to the task owner.
AI changes require human confirmation and are not automatically published. Review the result, test it where appropriate, and use the designer’s normal save and publish controls.
Tips for Better Results
Start with a specific business task.
“Create a supplier list for tracking contracts and renewal dates” provides more useful context than “Create a list.”
Name existing resources clearly.
If you want to use an existing Data List or approval form, include its name and explain its role.
State important rules explicitly.
Specify required fields, status options, approval conditions, or scheduling requirements instead of leaving them implied.
Review the plan before creation.
Use the Implementation Plan to resolve missing or incorrect details before generating the resource.
Make focused follow-up requests.
When refining a resource, describe the particular change you need. In the designer, use Add to chat to identify the relevant control or node.
Frequently Asked Questions
Does clicking Start with AI immediately create the resource?
No. AI first prepares an Implementation Plan. Review and confirm the plan before the resource is created.
Can I change the proposal before creation?
Yes. Explain the changes in the conversation and review the revised plan before confirming it.
Can I continue the conversation after creation?
Yes. Use Ask AI to reopen the original saved conversation and continue making changes.
What is the difference between Ask AI and Build with AI?
Ask AI returns you to the original saved conversation after resource creation.
Build with AI lets you refine supported resources inside the designer. You can use Add to chat to give AI context about a selected control or workflow node.
Are AI changes automatically published?
No. Changes made through Build with AI require human confirmation and do not automatically publish the resource.