AI Assistant
Pega AI Assistant™ is a conversational development tool in Infinity Studio that enables users to interact with Pega Platform™ through natural language. Instead of relying only on traditional form-based configuration, AI Assistant introduces a conversational layer that helps transform ideas into working application components.
By combining natural language interaction with structured and reviewable design, AI Assistant accelerates development, improves your understanding of applications, and supports troubleshooting, testing, and planning activities. It works alongside developers by responding with clarifications, suggestions, or direct updates that are visible immediately in the application.
AI Assistant also operates within a controlled environment by isolating its changes in a dedicated branch, ensuring that generated updates remain reviewable and safe.
For an overview of the AI Assistant, watch the following video:
Video transcript
AI Assistant is a conversational interface for communicating with Pega Platform that can help you transform natural language requests into design choices. AI Assistant supports GitHub Copilot, OpenAI, or Anthropic as an AI provider, and enables you to choose between several AI models that vary in speed, token cost, and work style. AI Assistant enables you to transform the initial blueprint into a functional application through conversation.
As you guide it through your goals and requests, it responds with information, clarifying questions, or direct changes to your application. For example, you can ask the AI Assistant to prepare service level agreements for you in bulk. The assistant immediately understands your application, its components, and the logic behind them.
To ensure a transparent development process, AI Assistant isolates all changes in a unique branch which constrains AI-generated edits to a safe workspace. When you review AI Assistant suggestions for accuracy, you can then merge them into your project.
Plan, Build, and Question modes
AI Assistant organizes its work through selectable modes that determine how it approaches tasks.
Plan mode
Plan mode focuses on analyzing requirements and defining a structured approach before making changes. In this mode, the assistant reviews the application context, asks clarifying questions, and proposes a design plan.
Build mode
Build mode runs approved plans by creating or updating application components. This includes generating Rules, configuring workflows, and implementing logic based on user input. Progress is displayed in the task list, allowing users to track completed and ongoing work.
Question mode
Question mode supports research and understanding by enabling users to ask questions about the application or about Pega Platform concepts. It retrieves context and explains logic, helping users learn and troubleshoot effectively.
Together, these modes enable a structured workflow that separates planning, implementation, and investigation while maintaining a conversational interface.
Common use cases
AI Assistant supports a wide range of development activities. The following examples illustrate how it can improve efficiency and productivity:
Planning and task management
- View open work items and organize them by priority or status
- Understand current development progress and next steps
Knowledge and research
- Retrieve definitions and explanations of Peg Platform concepts
- Provide contextual information about application elements
Application design
- Build workflows by describing requirements in natural language
- Generate a plan for a new Case Type and create it after approval
Business logic creation
- Suggest logic, such as Data Transforms, based on requirements
- Create draft Rules that can be reviewed and refined
Data Modeling
- Create Data Objects based on business needs
- Add or modify fields within Data Models
Integration and APIs
- Assist with connecting to external services
- Analyze OpenAPI specifications and map data
Application analysis and explanation
- Explain how flows and decisions work
- Provide visual summaries of application logic
AI Assistant adapts dynamically based on context, prior interactions, and selected model, encouraging iterative and conversational development.
Supported Rules that you can modify
The following supported Rule Types define the scope of what AI Assistant can create and modify during development:
- Activities
- Application Security
- Attachment Categories
- Case Types
- Class Metadata
- Class Rules
- Connect REST
- Correspondence Types
- Correspondences
- Data Pages
- Data Transforms
- Decision Tables
- Edit Validates
- Field Values
- Flow Actions
- Flows
- GenAI Agents
- GenAI Connects
- GenAI Tools
- Job Schedulers
- Notifications
- Pega Unit Test Cases
- Properties
- Queue Processors
- Report Definitions
- Service Level Agreements
- Service MCPs
- Validates
- Views
- When Rules
Check you knowledge with the following interaction:
This Topic is available in the following Module:
Want to help us improve this content?