Embedding
Published Updated
What is Embedding?
A numerical representation of text, images, or audio that captures the underlying semantic meaning of the data, allowing computers to understand relationships between concepts.
An embedding transforms text into an array of floating-point numbers (a vector). Words or sentences with similar meanings will have vectors that are mathematically close to each other in a multi-dimensional space. This allows databases to execute "semantic searches," finding relevant information based on meaning rather than exact keyword matches.
How it relates to Orbit
Memtrace, the code intelligence built into Orbit, offers semantic code search: it finds code by meaning rather than by exact text. See Memtrace Code Intelligence for details.
When you would use it
You use embeddings whenever you build a system that needs to search through large datasets to find contextually relevant information to feed to an LLM.
Common operations
- Converting enterprise documents into vectors for storage in a vector database.
- Executing a semantic search to find relevant context for a RAG pipeline.