Semantic Search¶

Semantic search uses AI to understand the meaning behind your questions, not just the keywords. Ask natural language questions and get relevant results based on context.

How It Works¶

When you ask a question in natural language, Outsprint converts your query into a vector embedding and searches across all your CRM data using Pinecone. This finds results based on meaning and context rather than exact keyword matches.

For example, searching for "budget concerns" finds notes and emails that discuss pricing, cost objections, or budget reviews -- even if they never use the exact phrase "budget concerns."

When to Use Semantic Search¶

Semantic search works best for:

Open-ended questions -- "Who did I talk to about pricing last week?"
Concept searches -- "Contacts interested in enterprise features"
Similarity searches -- "Find contacts similar to our best customers"
Activity recall -- "What happened with Acme last quarter?"

You

Which contacts mentioned budget concerns?

Outsprint AI

I found 5 contacts with notes or emails referencing budget discussions. Sarah Chen and Mike Park had the most recent mentions.

Hybrid Search¶

For most queries, Outsprint uses both full-text and semantic search together:

Full-text search finds exact keyword matches
Semantic search finds contextually relevant results
Results are merged and deduplicated
Permission filtering is applied
The best results are returned

This hybrid approach ensures you get both precise matches and relevant discoveries.

What Gets Embedded¶

Outsprint creates searchable embeddings for:

Contacts, companies, deals, and tickets
Notes and activity logs
Emails (synced and sent)
Knowledge base articles

Embeddings are automatically updated when records change.

Note

Semantic search results respect the same permissions as regular search. You only see records you have access to.

Pro Tip

Ask questions the way you would ask a colleague. Outsprint understands natural language, so you don't need to use special syntax.