Welcome to our Community Week MVP Insights series! Join ServiceNow MVP Ravi Gaurav as he discusses an idea he has around ServiceNow Agentic AI and self-managing catalogs.
Introduction
Most service catalogs in ServiceNow are static. They rely on manual updates from admins and often become cluttered with duplicate or outdated items. Over time, this makes the user experience harder, not easier.
I started thinking about what it would take for a catalog to manage itself. What if it could learn from how people use it and improve on its own? That idea led to what I call the “self-evolving service catalog,” powered by Agentic AI.
The Idea
The goal is simple: make the catalog smarter over time. Instead of waiting for someone to clean up old items or create new ones, AI can observe how users interact with the catalog and take action automatically.
Here’s how it would work:
- Track which catalog items are requested most often, and which are never used.
- Notice repeated manual requests that could be turned into new catalog items.
- Suggest updates to forms, categories, or keywords based on how users search.
- Flag outdated items or create draft versions of new ones for admin review.
The catalog doesn’t publish changes on its own. It recommends, learns, and proposes improvements. Admins still stay in control.
How It Works
The system connects four main parts:
1. Data source – It reviews records from requests, catalogs, and Virtual Agent conversations.
2. AI engine – A Script Include called AgenticAICatalog looks at user text and finds intent or similar patterns.
3. Automation layer – Flow Designer updates item data or generates draft entries.
4. Governance – Admins approve suggestions before anything goes live.
This balance keeps the setup intelligent but accountable.
Example Scenario
A user opens the portal and types “Need a dual-monitor laptop.” The AI checks the catalog. If it finds nothing similar, it automatically creates a draft for a new catalog item and notifies the admin.
The admin reviews the AI suggestion, edits if needed, and approves it. The next time someone makes a similar request, the catalog already has a ready-to-use item.
Why It Matters
This approach shifts ServiceNow from being a platform that reacts to requests to one that learns from them.
A self-evolving catalog could:
- Cut catalog maintenance work by more than half
- Improve accuracy in user searches
- Keep services up to date without constant admin intervention
- Help organizations respond faster to changing needs.
Admins still decide what gets published, but AI handles the routine discovery and cleanup.
What Comes Next
In future iterations, this setup can:
- Integrate Now Assist for richer recommendations
- Use feedback from fulfilled requests to fine-tune AI suggestions
- Support multiple languages for global catalog use
Final Thought
AI in ServiceNow isn’t just about automation. It’s about building systems that learn and adapt from experience. The self-evolving catalog is one small step toward a platform that helps teams work smarter every day.
Questions for Ravi? Leave them in the comments! Mark this post as helpful if you found it as such!
