Embedding: An embedding is a list of numbers that represents the meaning of a piece of text, an image, or other content, so that items with similar meaning end up with similar numbers. Embeddings are what let AI systems search by meaning instead of exact words.
Computers cannot compare meanings directly, so an embedding model converts content into a point in a high-dimensional space, typically a few hundred to a few thousand numbers. 'Invoice past due' and 'unpaid bill' land close together; 'quarterly offsite agenda' lands far away. Distance between points stands in for difference in meaning.
This is the machinery behind semantic search, recommendation, duplicate detection, and the retrieval step in retrieval-augmented generation. Your documents are embedded once and stored; a question is embedded on the fly and matched against them.
Embeddings capture meaning as the model learned it, so they inherit its blind spots. Highly specialized jargon, product codes, and names may not embed well without extra handling, which is why good systems combine semantic matching with plain keyword search.
Example at work
A customer support lead wants to find every past ticket about a recurring shipping-label bug, but customers describe it a dozen different ways. Instead of guessing keywords, the team embeds all tickets and searches for ones near a well-written description of the bug, surfacing hundreds that a keyword search missed.
Why it matters
You will rarely touch an embedding directly, but every AI search and 'chat with your data' feature depends on them. Knowing that matching happens by meaning explains both the pleasant surprises (finding what you meant) and the misses (a product code that never gets found).
Related terms
- Vector databaseA vector database stores embeddings, the numeric representations of text, images, or other content, and finds the items closest in meaning to a query quickly. It is the storage layer behind semantic search and most document-aware AI assistants.
- Semantic searchSemantic search finds results based on the meaning of a query rather than exact keyword matches, so a search for 'unpaid invoices' also returns documents about 'overdue bills.' It uses embeddings to compare meaning.
- Retrieval-augmented generation (RAG)Retrieval-augmented generation (RAG) is a technique in which an AI system first searches a set of documents for relevant passages and then gives those passages to a language model to answer from. It lets an assistant use your own, current information.
- Neural networkA neural network is a computing system made of layers of simple connected units that pass numbers to each other, with connection strengths (weights) adjusted during training until the network produces useful output. It is the building block of modern AI.