Enabling AI features in servicenow ITSM module

Murali7
Kilo Contributor

3 weeks ago

Iam working  as a Business Analyst, Could you please guide me how to gather requirements for enabling AI feature in ITSM application

Labels:
0 Helpfuls
  • 785 Views
3 REPLIES 3

Aditya_hublikar
Giga Sage

3 weeks ago

Hello @Murali7 ,

As a Business Analyst, the first step in enabling AI in ServiceNow ITSM is to understand the client's business objectives, existing processes, and challenges. Conduct discussions with stakeholders, such as ITSM managers, service desk agents, and technical teams, to identify suitable AI use cases, including incident summarization, resolution notes generation, and knowledge assistance. Gather functional and non-functional requirements related to functionality, user access, security, data privacy, and performance. Collaborate with the technical team to validate licensing, compatibility, and implementation feasibility. Finally, document the requirements in a BRD, user stories, and acceptance criteria, define measurable success metrics .

 

If you want technical guidence for Servcienow NowAssist then you can refer servicenow course where will get all required details :

https://learning.servicenow.com/lxp/en/it-service-management/now-assist-for-it-service-management-it...

 

Community post :

https://www.servicenow.com/community/servicenow-otto-forum/gf-8-getting-started-with-now-assist/td-p...

 

 

If this helps you then mark it as helpful and accept as solution .

Regards,

Aditya

0 Helpfuls

BhavaniYamsani
Mega Guru

3 weeks ago

Hey @Murali7 ,

As a Business Analyst, I would approach enabling AI in ServiceNow ITSM in two parts:

1. Functional requirements — Business perspective

Discuss the following with the ITSM process owner and service desk agents:

  • Business objectives: What problems are we trying to solve?

  • Use cases: Do we need incident summarization, resolution notes generation, knowledge assistance, or change risk explanation?

  • User interaction: Should AI run manually or automatically?

  • Expected output: What information should AI generate, and where should it appear?

  • Human approval: Should agents review the AI output before saving or acting on it?

  • Success criteria: How will we measure improvement, such as reduced handling time or improved resolution quality?

Deliverable: Business requirements, user stories, process flows, and acceptance criteria.

2. Technical requirements — Platform perspective

Work with the ServiceNow Architect and Developer to validate:

  • Platform compatibility: Confirm the ServiceNow release, patch, and required application dependencies.

  • Licensing: Verify whether Now Assist for ITSM and the selected capabilities are included in the organization's license.

  • Configuration: Identify the required plugins and OOTB skills.

  • Data requirements: Confirm the tables, fields, related records, and knowledge sources needed by the AI skill.

  • Security: Define user roles, ACL requirements, data privacy, and access restrictions.

  • Customization: Determine whether OOTB functionality meets the requirement or a custom skill is needed.

  • Testing: Define UAT scenarios, output quality checks, error handling, and deployment criteria.

Deliverable: Technical feasibility assessment, dependencies, security requirements, and test scenarios.

3. Recommended implementation sequence

  1. Identify and prioritize one business use case.

  2. Document the current process and desired outcome.

  3. Prepare functional requirements and acceptance criteria.

  4. Validate licensing, platform compatibility, data, and security with the technical team.

  5. Configure the selected Now Assist skill in a sub-production instance.

  6. Conduct UAT with representative users and measure the results.

  7. Obtain business sign-off before production deployment.

For implementation guidance, refer to the official Configure Now Assist for ITSM documentation.

Thankyou

Yamsani Bhavani

ServiceNow Developer - SecOps, GRC, Custom Applications, Now assist

0 Helpfuls

yashkamde
Giga Sage

3 weeks ago

Hello @Murali7 ,

 

To gather the requirements, you will first need to know what are you implementing Generative AI (Now Assist), AI Agents, Agentic AI.

What specific metric is your priority?

 

Steps for requirement gathering process :

  • Identify specific use cases, focus on areas like automatic ticket routing, incident summarization, or chat deflection. 
  • Assess data quality and verify you have clean, sufficient historical ticket data to train the models effectively.
  • Define user transition paths like map out exactly when the AI should step back and hand the conversation over to a live human agent.
  • Define thresholds, set the minimum accuracy level (e.g. 80%) required before letting the AI automate actions like field population or categorization.
  • Also confirm your ServiceNow licensing tier (Pro/Enterprise/Now Assist) and align data privacy guidelines with your security team.

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

0 Helpfuls