Retrieval & search
Semantic search over your documents
Semantic search finds documents by meaning rather than exact keywords: it embeds both your query and your documents into vectors, then ranks by similarity, so "could not sleep and lost interest in things" matches a note about insomnia and anhedonia even with no shared words. The demo below searches a transcript library by meaning, fully in your browser.
Try it live, no sign-up
Similar Transcript Finder
Semantic search over consultations: meaning, not keywords.
Why this matters now
Keyword search misses paraphrases, synonyms, and other languages, which is exactly how real users phrase things. Semantic search is also the retrieval primitive under every RAG system, so getting it right is the foundation for trustworthy question-answering over your own content.
How the reliable version works
- 1Embed each document into a vector that captures its meaning.
- 2Embed the incoming query the same way.
- 3Rank documents by cosine similarity to the query vector.
- 4Optionally rerank the top candidates for precision.
- 5Return results by meaning, across paraphrases and languages.
Common questions
How is semantic search different from keyword search?
Keyword search matches the words you typed. Semantic search matches the meaning, so it finds relevant content that uses different words, synonyms, or another language. For natural-language queries it is far more forgiving and far more useful.
Does it work across languages?
Yes, with a multilingual embedding model. The same query can match relevant documents written in other languages, because both are mapped into the same meaning space. I have shipped this for multilingual clinical search.
Do I need an API key to try it?
No. The demo runs the embedding model on-device in your browser, so you can search by meaning with no key and nothing uploaded.
Related demos, work and services
Want this built properly for your product?
Tell me what you are building and where it is stuck. I reply within one business day with an honest read on whether I can help.
Send a project inquiry