Vector Representation (Embedding)
Definition
An embedding is a numerical representation of text, images or other inputs as a vector in a high-dimensional space. Inputs that are similar in content lie close to each other in vector space, which enables semantic search and comparison.
Meaning in a Business Context
Embeddings are the technical foundation of semantic search and therefore the prerequisite for retrieval-augmented generation. A typical RAG workflow works like this: all company documents are converted into embeddings and stored in a vector database. When a user asks a question, the question is also converted into a vector, the most similar documents are retrieved and provided to the language model as context. Similar data protection considerations apply to the choice of embedding models as to language models: are the contents transmitted to an external provider, or does the embedding run locally? Open-source models such as BGE or Multilingual-E5 allow fully on-premises operation.