Pega Infinity Studio and Pega AI Assistant

Pega Infinity Studio supports a development process that progresses from initial prototyping in Pega Blueprint® to AI-assisted development and manual refinement.

The structure and intent of your Blueprint continue into Pega Infinity Studio as you build, configure, and refine your application by using the latest Pega GenAI™ tools. After you import a Blueprint into a Pega Platform instance, Pega GenAI creates a list of Tasks that connect the design and implementation phases. Each Task includes a plan that describes the steps required to complete the work. You can complete the Tasks manually or by using AI Assistant. 

This process does not replace developers. AI Assistant reduces repetitive work and gives your team more time to make high-value design and implementation decisions. Infinity Studio combines the tools and features of App Studio and Dev Studio. This unified environment reduces context switching between development environments. 

The navigation menu on the Infinity Studio Home page reflects the Blueprint user experience. The areas that you define in Blueprint, such as workflows, data and integrations, and Personas, appear in the same structure in Infinity Studio. 

The following figure compares the navigation menu in Blueprint with the navigation menu on the Infinity Studio Home page:

The Infinity Studio navigation menu includes an additional node, Implementation Plan. The Implementation Plan stores pending Tasks and configurations that originate in Blueprint. Your design therefore enters Infinity Studio as a plan of record instead of a document that a developer must interpret and plan again. By tracking Tasks in the Implementation Plan, you can better understand the development effort and plan, assign, and complete work more efficiently. 

For more information about navigation in Infinity Studio, see Exploring Infinity Studio. 

Pega AI Assistant adds a conversational layer to Pega Infinity Studio. AI Assistant combines natural-language interaction with a structured, reviewable design process. It can accelerate development, help you understand applications, and assist with troubleshooting, testing, and planning. AI Assistant can provide clarifying questions, suggestions, or direct updates that are immediately visible in the application. 

AI Assistant can help you complete the following development activities:

• Review Implementation Plan Tasks 
• Explain existing applications and Rules 
• Generate business logic and associated Rules 
• Troubleshoot application issues 
• Generate and run automated tests

AI Assistant applies the following characteristics to each development session: 

Context awareness 
AI Assistant understands the context of your application, including its components and underlying logic. During a conversation, you can direct AI Assistant to a specific context, such as a Rule, an attachment, or a Task. 

Built-in best practices 
AI Assistant uses built-in skills and workflows that Pega experts create. These skills apply Pega best practices and help AI Assistant produce consistent, repeatable results. 

User-scoped permissions 
AI Assistant operates under the security permissions of the person who directs the session. Its access rights depend on the Access Group of that person.  

Branch-based development 
AI Assistant isolates development in Branches. Branch isolation limits the effect of AI-generated changes on the rest of the application. You review AI-generated changes before you merge them into a work Branch. 

To open AI Assistant, in the lower-left corner of the Infinity Studio navigation pane, click the AI Assistant icon. The interactive AI Assistant pane remains open while you work. The following figure shows the AI Assistant pane during development:

The AI Assistant pane contains the following areas:

1. General settings 

Open a new chat, switch to full-screen mode, and open the Actions menu. The Actions menu provides access to chat history, provider settings, performance reports, feedback forms, and information about the latest changes to AI Assistant. 

2. Responses

Displays the answers and follow-up questions that help AI Assistant understand your intent. 

3. Context

Lists the sources from which AI Assistant collects the information that it requires to complete its tasks.

4. Input

Enter requests, switch to voice commands, add attachments, or add a Rule to the context by clicking the Rule directly in the UI.

5. Mode

Switch the focus of AI Assistant between planning your work and implementing your plan.

6. Model

Select an AI model. Models vary in speed, token cost, and work style. 

Plan and Build modes

AI Assistant has two modes that separate design decisions from implementation. The following table describes each AI Assistant mode: 

When you first open the AI Assistant pane, Infinity Studio prompts you to select an AI provider. You can connect AI Assistant to the following providers: 

• GitHub Copilot
• Anthropic
• OpenAI
• A Pega-managed model in a Pega Cloud® environment, with no external setup or token cost

 For more information, see Connecting AI Assistant to AI providers. 

AI Assistant can help you understand an application by explaining an open Rule. On the Rule form, in the upper-right corner, click Actions > AI Assistant. Infinity Studio opens the AI Assistant pane and creates a request for the selected Rule. AI Assistant provides the following commands on a Rule form: 

• The Explain command provides a high-level summary of the Rule.
• The Review command reviews the Rule and identified potential improvements.  

You can also access these commands directly from the AI Assistant pane by entering a forward slash (/). The following figure shows the AI Assistant option in the Actions menu of a Data Page Rule: 

AI-assisted development in Infinity Studio relies on several architectural components that work together to process user requests, access application metadata, invoke authoring tools, and interact with the selected large language model (LLM) provider. 

When  you submit a request through AI Assistant, a set of services uses application context, built-in Pega skills, and the selected AI provider to process the request. Depending on the request, AI Assistant can analyze application artifacts, propose implementation approaches, generate tests, and create or modify supported Rules. The GenAI Component provides application schemas and authoring capabilities that support these operations within the authorized context. After the required model interactions and authoring activities are complete, the results return to AI Assistant in Infinity Studio. 

The following figure shows the interaction between the components that support AI-assisted development:

Pega AI Assistant applies built-in best practices for agentic development to keep AI-generated changes reviewable, traceable, and under your control. 

Branch isolation

AI Assistant isolates AI-generated changes in a unique Branch. The Branch isolates AI-generated changes from other development work until you review and merge them. This approach helps maintain visibility and control over AI-assisted updates before they become part of the application. When you work from a Implementation Plan Task, AI Assistant uses the Task ID as the Branch ID if you have not set a Branch preference. If a Task-based Branch ID is unavailable, AI Assistant creates a semantically relevant Branch ID that describes the work. 

Tracking AI-assisted changes

AI Assistant provides visibility into Rules that it creates or modifies. In the Branch view, the AI involvement column identifies the level of AI involvement for each Rule: 

• Generated indicates a Rule created by AI Assistant without subsequent user modifications. 
• Assisted indicates a Rule that includes AI-generated changes and was also modified by a user.
• None indicates that AI Assistant was not involved in the Rule change. 

The Branch view therefore provides a consolidated way to identify AI involvement across the Rules in the Branch. 

AI involvement is also reflected in the Rule history. History memos indicate when a Rule was created or modified through AI-assisted authoring, providing a record of AI involvement alongside other Rule history entries. 

Together, the Branch view and Rule history provide visibility into AI-assisted changes at different levels. The Branch view shows the AI involvement classification, while Rule history records AI-assisted authoring activity for an individual Rule. 

Performance reporting

The AI Assistant Performance Report provides transparency and auditability by exposing key execution details behind AI-generated outcomes. Information captured includes token consumption, model selection, execution time, and tool interactions. You can export and share the report to support analysis, troubleshooting, and governance activities. 

To open the performance report, in the AI Assistant pane, click Actions > Performance report.

 Check your knowledge with the following interaction: