Skip to content
Orbit Docs

Retrieval memory for long-context work

← Learning Atlas · Path 5 of 22

6 ordered steps, from Embedding to Token.

AI providers and modelsHostinginfrastructuredatabases

  1. Step 1: Embedding

    Embeddings come first because retrieval systems need a way to turn meaning into searchable vectors before they can recall anything useful.

  2. Step 2: RAG

    RAG uses those vectors to bring outside knowledge into a model session, making it the first complete pattern after embeddings.

  3. Step 3: Pinecone

    Pinecone gives the abstract retrieval pattern an external managed database where those vectors can actually live.

  4. Step 4: Chroma

    Chroma adds a local or developer-friendly alternative, helping learners compare managed retrieval with lighter-weight setups.

  5. Step 5: Context window

    Context window matters after retrieval because it explains the hard limit that retrieved material still has to fit inside.

  6. Step 6: Token

    Tokens are the unit that makes the context limit measurable, so they close the loop from search result to model input budget.