Hi , how to implement Pega chat with Gen AI Agent, we want to open Pega chat with customer where Customer can chat with GenAI and eventually it will create an interaction, please share a direct and simple approach, thank you
If by offline you mean the app must work without network connectivity, then GenAI itself cannot run offline because it requires a live model call. Offline in Pega means the app can keep working without connectivity and sync later, but GenAI needs a real-time model call, so that part has to be online.
So, if you mean true offline mode, then GenAI is not the right fit for the live conversation part.
The realistic design is to let Pega capture the chat messages offline, sync them when connectivity returns, and then continue the GenAI conversation online.
- Offline: capture the chat message locally or queue it.
- Online: once connectivity is back, send it to GenAI.
- Then: create the interaction in Pega based on the chat outcome.
If you need a direct customer chat experience with GenAI, that part has to stay online. Offline mode can help with draft capture, queueing, and later case creation, but not real-time AI responses
Hi thanks for the reply, I have removed the offline word, I meant no CSR will be available and customer will be able to chat and interact with GenAI to get the reply/answers, thanks
It would be a direct approach which can be separated this into two parts: chat experience and case creation.
Let the customer chat with Pega through a chatbot or digital messaging channel, and use a GenAI step behind the scenes to handle the conversation and understand the intent.
When the conversation reaches a clear business trigger, such as “create a service request” or “open an interaction,” have the bot hand off to a Pega case and create the interaction at that point. That keeps the chat natural while still using Pega for orchestration and case management.
don’t try to make GenAI create the interaction from the first message. Let it converse first, collect the details, and then create the interaction when the intent is clear.
Hello , You can use the agentic messaging channel to meet your requirements. This allows clients to interact with a GenAI agent to fulfill their requests.
We can configure the Agentic messaging feature from Pega 25.1.2. To see the agentic messaging channel in the channels, please update the Pega-Landing-AppView-CognitiveAI-Messaging.pyIsAvailable when rule to return true. For creating and configuration steps, please refer to Setting up an Agentic Messaging interface and a Web Messaging Channel | Pega Academy
Thank You.
Excellant, thanks but I dont see Agentic messaging and Agentic email option, can you guide me how to see those, also Is Agentic messaging different from the Digital messaging ? thanks
Yes Agentic messaging is different from the Digital messaging.
To create agentic messaging channel, please update the Pega-Landing-AppView-CognitiveAI-Messaging.pyIsAvailable when rule to return true to see the agentic messaging channel option in channels landing page and its applicable from the Pega CS 25.1.2 release.
Thank you.
thank you again and may I ask, which Agent have you used to integrate with chat flow in agentic messaging ? like in the channel, which OOTB agent we should use?
1 Like
As of now, we have not shipped any out‑of‑the‑box (OOTB) agent rules for the agentic messaging channel. You need to create your own GenAI Agent rule from App Studio → AI Designer based on your specific requirements.
To learn more about creating and configuring GenAI Agent rules, please refer to the following documentation:
Automating work with GenAI agents
Thank You.
Hi thank you, Does it mean we need to write our own prompt ? thank you