How to auto-detect the existing known errors when raising new problems?

[sattar3]

Hello All,

Problem management process owner wanted the below requirement:

As a process owner: I want to ensure that we minimise duplicate problem investigations and consequently improve the efficiency and effectiveness of our problem management capabilities. The known error database must be cross-referenced with the known error database (KEDB), using AI NLP and pattern matching. The check should occur during logging, classifying and prioritising a problem, and before starting the RCA investigation. Ideally it can also be repeated during the ongoing RCA with the discovery of new information.

Does anybody worked this kind of stories? 

Thanks,

Sattar

[kaushal_snow]

@sattar3 ,

You can use ServiceNow Predictive Intelligence Similarity Solution to automatically cross reference a new problem record against your existing Known Error Database (KEDB) by creating a similarity model that uses fields such as short description and description, training it on prior known error/ problem records to surface high confidence matches whenever a new problem is logged (or when new data is added during the RCA phase), thereby helping you avoid duplicate investigations by prompting the analyst with potential existing errors before launching a full root-cause investigation......

If you found my response helpful, please mark it as ‘Accept as Solution’ and ‘Helpful’. This helps other community members find the right answer more easily and supports the community.

[sattar3]

Hi @kaushal_snow , thanks for replying.

But we don't have 10k records and we are not allowed to change the system property.

Is there any other way without using Predictive Intelligence to achieve this task?

Thanks,

Sattar

[Dr Atul G- LNG]

Hii @sattar3 

Agree with @kaushal_snow 

PI is the best option to use here. We have implemented this with our customer base when using PI, and it worked well. PI offers multiple formats, such as classification and grouping, where you can set patterns as required. Since it is a licensed feature, you’ll need to check availability. Otherwise, there’s no direct way, though it might exist in the repository—try searching with relevant keywords