Category · AI

Embedding

Vector representation of a piece of text or media.

A dense vector (768-3072 dimensions) produced by a model (OpenAI text-embedding-3, Cohere, Voyage). Used to measure semantic similarity (cosine) between documents in a vector DB. Foundation of retrieval in a RAG pipeline.

// In action with our clients

Client cases

/10European bank: hybrid RAG, inference cost cut tenfold.

IARAG in production: from €10,000 to €900 per month.

// See also

RAG (Retrieval-Augmented Generation)
AI architecture that combines document retrieval with LLM generation.
Vector DB
Database optimized for ANN vector queries.

Want us to apply this for you?

Talk to an architect