Semantic Search
Find content by meaning, not just keywords — powered by AI embeddings.
Semantic search goes beyond matching words. It understands the meaning behind your content and finds records that are conceptually related to what you are looking for.
How It Differs from Text Search
With regular text search, you need to use the right keywords. Semantic search removes that limitation. For example, searching for "error handling" can surface a record about exception handling, retry strategies, or debugging patterns — even if none of those records contain the exact phrase "error handling."
This works because Goodterm creates vector embeddings of your content. These embeddings capture the meaning of your text in a way that allows the AI to compare concepts, not just words.
Where Semantic Search Is Used
You do not need to activate semantic search manually — it works behind the scenes in the places where it matters most:
- AI Chat — When you send a message, the AI automatically uses semantic search to find relevant records in your knowledge base. This is how the AI knows which of your records to draw from.
- Context panel — The related records shown in the chat sidebar are found through semantic search.
Getting the Best Results
Semantic search works best when your records have substantial, well-written content. A few tips:
- Write naturally — the AI understands prose better than cryptic abbreviations
- Add context and explanations to your records, not just bullet points
- The more records you have, the more connections the AI can discover