Queries on AI Workflow/Usecase

Vasu ch
Kilo Sage

Hello,

 

I have been exploring Now Assist and AI Workflows, and I have a few questions. I’m hoping to get them answered here.

 

  1. I have created my AI Workflow with all the instructions and agents. How can I train this workflow to make it more efficient?

  2. While testing the AI Workflow, I sometimes observed that the AI hallucinates and provides incorrect outputs. Is there a way to provide feedback in such scenarios to let the AI know that the output is incorrect?

  3. I see that we can allow our AI Agents to learn from past executions. Can this training data be moved across instances?

 

Regards,

Vasu

1 ACCEPTED SOLUTION

rpriyadarshy
Tera Guru

@Vasu ch here is my 2 cents on this.

 

  1. I have created my AI Workflow with all the instructions and agents. How can I train this workflow to make it more efficient?---->You do not “train” an AI Workflow in the traditional ML sense. Efficiency is improved through design-time optimization, data grounding, and instruction refinement. Since LLM is owned and controlled by Servicenow so no training we can impart. But based on Ai workflow output you can mark is thums up or down (sys_gen_ai_log_metadata). This goes as a feedback on LLM response to Servicenow Team.  You can improve efficiency - Use Case Instruction refinement, Agent Instructions,Roles , Scope of work ,  lesser no of agents to reduce orchestration complexity

 

  1. While testing the AI Workflow, I sometimes observed that the AI hallucinates and provides incorrect outputs. Is there a way to provide feedback in such scenarios to let the AI know that the output is incorrect?--> There is currently no direct user-feedback loop that teaches the LLM that an output was wrong. You can refine prompts, instructions, and grounding data based on incorrect outputs to reduce hallucination.

Some URLs to refer

https://www.servicenow.com/community/now-assist-articles/ai-agents-faq-and-troubleshooting/ta-p/3200...

https://inmorphis.com/insights/blogs/reducing-genai-hallucinations-in-servicenow-a-rag-based-governa...

 

  1. I see that we can allow our AI Agents to learn from past executions. Can this training data be moved across instances?--> NO

AI Agent Practical Implementation: Lessons from th... - ServiceNow Community

 

Regards

RP

 

View solution in original post

6 REPLIES 6

Dr Atul G- LNG
Tera Patron

Hi @Vasu ch You can go to the Best Practice site, where you will find some useful discussions about AI. You can review them, and they might be helpful.

*************************************************************************************************************
Regards
Dr. Atul G. - Learn N Grow Together
ServiceNow Techno - Functional Trainer
LinkedIn: https://www.linkedin.com/in/dratulgrover
YouTube: https://www.youtube.com/@LearnNGrowTogetherwithAtulG
Topmate: https://topmate.io/dratulgrover [ Connect for 1-1 Session]

****************************************************************************************************************

Sure Atul. Will do.

 

Regards,

Vasu