Pega is making application development faster with its built-in AI Assistant. Inside Pega Infinity Studio, you can now power that assistant with your own AI coding tool — like GitHub Copilot — instead of relying only on Pega’s built-in model. Once connected, you can simply describe what you want in plain English, and the Pega AI Assistant will update your case’s Stages and Steps directly, following Pega’s own best practices behind the scenes. This article walks through connecting GitHub Copilot to Pega, exploring the assistant’s toolbar, and then using it to add a Step to a Stage. Let’s get started.
Getting to Know the AI Assistant Toolbar
Before diving into configuration, it helps to know what’s available in the AI Assistant chat panel:
- Build / Plan mode – A dropdown lets you choose how the assistant works. Build makes changes directly — editing Flow rules, adding Steps, updating Case Types, and so on. Plan is a read-only mode where the assistant lays out a proposed set of changes for you to review before anything is actually applied.
- Dictate – A microphone icon lets you speak your request instead of typing it, useful for quick, hands-free instructions.
- Attachments – A paperclip icon lets you attach files, such as documents or images, to give the assistant extra context for the task.
- Select icon – A target-style icon lets you explicitly select a Stage, Step, or other object as context, so the assistant knows exactly what part of the case you mean instead of guessing from the conversation alone.
- Model selector – A dropdown (for example, showing Claude Sonnet 5) lets you choose which underlying model powers the response, independent of which provider is connected.
- Full screen – An expand icon at the top of the panel enlarges the AI Assistant into a full-screen view, handy for longer conversations or detailed change summaries.
Configuration Steps
- Open the AI ASSISTANT panel on the left side of Pega Infinity Studio.
- Go to Connect an AI provider (or Switch AI Provider if one is already set up).
- Under GitHub Copilot, click Sign in.
- In the Sign in to GitHub Copilot dialog, copy the device code shown (e.g.,
8523-DA35). - Click Open Verification Page to go to GitHub’s authorization page in a new tab.
- Paste the device code on GitHub and authorize the connection.
- Return to Pega and click Check Status.
- Confirm the “You’re All Set — GitHub Copilot is now connected to your AI Assistant account” message.
- Optionally, pick the underlying model from the dropdown at the bottom of the chat box (e.g., Claude Sonnet 5).
- Open a Case Type (e.g., Payment Processing) and select Build mode in the AI Assistant chat.
- Tag the Case Type for context, then type your request in plain English, for example: “add a new Step review documents after gather documents in payment intake stage.”
- Review the assistant’s summary of the updated Flow and the new Step sequence.
Demo: Click here
Closing Thoughts
Connecting GitHub Copilot to Pega takes just a sign-in and a device code — no manual API wiring required. And once connected, the Pega AI Assistant does more than fill in a form: it edits the real Flow rule, explains exactly what changed, and still pauses for your approval before anything is finalized. That balance of flexibility and governance is what makes this feature genuinely useful for Pega teams, not just a demo trick. Try it on a sandbox case type and see how it fits into your own build process.