Search options in BMC Helix ITSM

 

BMC Helix ITSM supports searching across multiple record types and resources. It provides the following search options tailored to different user needs:

• Global Search: It is a keyword-based search across tickets (incidents, change, problems), assets, knowledge articles, and people. You can use this to find  items by title, ID, or terms across multiple sources at once.
• Agentic Chat: It is a conversational, semantic search to find the most relevant answers from knowledge articles, incidents, and other enterprise data. The system evaluates the meaning of a user question and retrieves answers with natural language summaries and linked sources.

How Agentic Chat semantic search works

Agentic Chat powered by BMC HelixGPT uses Large Language Models (LLMs) and vector‑based indexing to understand the intent behind a user query. When a user enters a question, BMC HelixGPT processes it through the following steps:

What differentiates semantic search from keyword search

BMC HelixGPT semantic search	Traditional keyword search
Matches concepts and intent	Matches exact words
Best when asking questions or troubleshooting	Best when searching for specific IDs or exact phrasing
Uses context from query and article content	Cannot interpret context
Returns a synthesized answer with supporting links	Returns a list of results

When to use each search option

Agentic Chat	Global Search
Use when you want: Explanations Troubleshooting steps Summaries of information Natural‑language answers	Use when you want: Exact items (tickets, knowledge articles, assets) Searches by ID (INC12345, KA‑000001)

Fields considered by BMC HelixGPT per use case

BMC HelixGPT processes different record fields depending on the use case. The amount and type of content included from these fields directly influences the information the model must analyze.

The following table summarizes which fields are considered for each use case.

Use case	Fields considered (affects processing)
Knowledge article-based questions	Title Problem Environment Symptoms Resolution Step‑by‑step instructions Tags, keywords, synonyms, acronyms Full article body text
Incident-based questions (Agent Assist)	Summary Description Work Notes / Activity Logs Resolution Categorization (Product / Operational) Priority, status Custom semantic‑attribute fields configured by admin
Change / Problem ticket questions	Summary Work notes Implementation (changes) Root cause / Known error (problems) Categorization Assignment and group fields
Chat‑based natural‑language questions	Retrieved knowledge article content (full body + metadata) Historical ticket data (summaries, notes, resolutions) Configured external sources Semantic tags and vector-indexed chunk
Autoresponder / email‑based AI processing	Email subject and body Incident fields mapped to autoresponder Prompts configured for summarization/extraction Dynamic content defined in templates

Where to go from here

Searching for tickets, assets, resources, and people in BMC Helix ITSM