Countless issues with Otto/Now Assist product

Suggy
Giga Sage

How many of you are happy with Otto/Now Assist? atleast for me, I see countless issues with this product.

Inconsistent, not stable, not fast, not reliable etc..

 

Many times I see the AI agents wont work, some or the other error.

 

Since today morning, when I am trying to create support case, I see this message:

Suggy_0-1791988976801.png

"Sorry, there was a problem on my side trying to complete this request. Try asking again later."

 

So in support portal only when we see this, I understood the reality of this products stability

 

4 REPLIES 4

yashkamde
Giga Sage

Hello @Suggy ,

 

I have faced similar issues multiple times while working with Now Assist and AI Agents.

 

However, from my experience, I feel some of these issues also arise when we are unable to provide sufficient context, configure the right tools, or define a proper process flow for the AI Agent.

When the required tools, instructions, and execution flow are properly configured with clear context, I have seen much more consistent and reliable responses.

 

If my response helped mark as helpful and accept the solution.

rpriyadarshy
Kilo Sage

@i also face this Issue intermittently and even i faced yesterday.

 

During the Issue i checked the gen AI Logs- I saw nothing was coming in Logs (Table Name-sys_generative_ai_log). It seems that during that time my ServiceNow Instance Gen AI Skill was not triggering because it was unable to commect to ServiceNow Now LLM behind the scene. Todays world there are so many components and all are in their Respective ServiceNow CLoud Infrastructure. At Application level we do not control the backend connectivity Issue much.

 

Regards

RP

How were you able to tell that the agent was not getting connected to the NOW LLM?

I want to be able to track possible causes when these things happens on my assistants

sanjaygupta
Mega Guru

Hi,

That message is generic and does not indicate what actually failed. The screenshot shows it on the support portal’s Case Creation flow, which runs on ServiceNow’s side. If that is the case, it cannot be resolved from your instance, and the available documentation does not list a status or fix at this time.

Community replies describe this message as typically a backend handshake failure or timeout between the instance and the ServiceNow LLM service (Generative AI Controller). A Community article lists it as a symptom of generative AI responses not loading, often following a partial plugin update.

If you see the same message on your own instance, the following steps come from the documentation and the Community:

  1. Open sys_generative_ai_log.list and look for errors at the time of the failure.
  2. Go to All > Now Assist Admin > Features and run the skill’s diagnostics. This requires the sn_nowassist_admin.nsa_admin role.
  3. Repair the Generative AI Controller plugin first. The article above reports this resolved the issue in several cases.
  4. Confirm AI Search is active at All > AI Search > AI Search Status.
  5. Update the parent Now Assist plugin so the dependent plugins update as well.
  6. If the issue persists, log a Now Support case and attach the log entries.

The Now Assist Troubleshooting app on the Store may also help with the intermittent behavior. It lets admins configure proactive error monitoring that alerts when execution errors cross a threshold.

If the error occurs on the Now Support portal, try the case form without Now Assist and raise the Now Assist failure through your account team.

Sources: