Executive Summary

The Semantic CMDB Search Tool is an AI-powered prototype that transforms ServiceNow PDI CMDB records into actionable insights. Users can ask questions in plain English to surface configuration items, relationships, metadata, and ownership details from the test dataset.

Inspired by my experience as an Application Portfolio Manager, this prototype solves the challenge of locating accurate information within large, distributed CMDB datasets. Traditional CMDB interfaces often require specialized knowledge and manual filtering, slowing discovery for IT teams and business stakeholders.

By bridging structured enterprise data with semantic AI search, the prototype modernizes workflows while keeping sensitive data private and secure.

Overview

Enterprise CMDBs can contain tens of thousands of applications, servers, integrations, and service components. Finding the right information quickly is difficult. This tool simplifies discovery by enabling natural language search across structured configuration data, making CMDB insights accessible to both technical and non-technical teams.

Key Features

  • Natural Language Search: Ask questions in plain English.
  • Semantic Understanding: AI interprets meaning, not keywords.
  • Private & Secure: All CMDB data and embeddings remain within your local PDI environment. Vectors are stored in a local Chroma DB, and queries are processed locally, ensuring no sensitive information leaves your system.
  • Fast Retrieval: Vector search returns results in seconds.
  • Scalable: Can ingest additional systems or data sources.

Target Users

  • Application Portfolio Managers needing fast visibility into systems and dependencies.
  • ServiceNow Admins wanting easier CMDB search and analysis.
  • IT Operations teams during incident response and audits.
  • Business stakeholders who need access without deep technical expertise.

Business Value

  • Efficiency: Faster CMDB discovery and analysis.
  • Accessibility: Makes complex data usable for all stakeholder groups.
  • Faster Decisions: Supports incident response, reviews, and audits.
  • Governance & Security: Maintains data integrity and privacy.

Technology Overview

  • OpenAI Embeddings: Power semantic understanding of user queries.
  • Chroma Vector DB: Stores embeddings and enables fast similarity search.
  • Python Backend: Orchestrates embedding generation and query routing.
  • Streamlit UI: Provides a clean, interactive interface.
  • ServiceNow PDI Integration: Retrieves realistic CMDB records for safe testing.

Note: The prototype connects to a ServiceNow Personal Developer Instance (PDI) to ensure safe, realistic testing without exposing live enterprise data.

MVP Scope

  • Connect to ServiceNow PDI to retrieve CMDB data.
  • Generate embeddings and store them in Chroma DB.
  • Process user questions through OpenAI embeddings.
  • Perform semantic similarity search.
  • Display results in Streamlit with key metadata fields.

Roadmap / Future Enhancements

  • Simplified deployment for broader adoption.
  • Role-based user access for secure multi-user environments.
  • Automated refresh of semantic data for dynamic CMDB updates.
  • Conversational, chatbot-style search interface.
  • Integration with additional IT and business data sources.