Does anyone know about long-term memory in servicenow ai agent?

Tae Kyung Lee · ‎04-15-2025

Hello.

I've found that there are two properties that enables long-term memory for AI Agents. (Link)

- sn_aia.ltm.enable_long_term_memory : Enables long-term memory for AI agents. All previous user interactions are used as context for the LLM.

- sn_aia.ltm.use_memory_for_ai_agent : Enables a long-term memory for AI agents. Previous user interactions are used in AI agent interactions.

Do you how and where they are stored in platform and can be accessed for ai agent?

warren_chan · ‎04-21-2025

Yes, this is what we refer to as long-term memory, and it will be stored in the AI Agent Memory [sn_aia_memory] table. You can test this by running the same Agentic AI test case over and over again, and responding to the conversations the same way each time. The long-term memory is user-specific, but this feature may improve over time.

View solution in original post

Tae Kyung Lee · 4 weeks ago

If you want to trace execution of AI Agent, you can see below tables.

- sn_aia_execution_plan

- sn_aia_execution_task

- sn_aia_tools_execution

For more detailed AI Agent schema model, refer here (Great article of @Nicola Attico)

warren_chan · ‎04-21-2025

You can try being explicit in your prompting. For example, if you want to test that long-term memory is being utilized, you can put a step in the prompt/instructions for the AI agent to "Check long term memory for any saved user devices. If saved user device found in long term memory, then ask user to verify that the device information is still current."