Compute document and query vectors with openai/text-embedding-3-large
Retrieve Top-K from a vector store (FAISS / Milvus / pgvector)
Re-rank with cohere/rerank-english-v3
Stitch the Top-N chunks into the prompt and let cohere/command-r-plus or Claude generate the final answer
Embedding
python
resp = client.embeddings.create(
model="openai/text-embedding-3-large",
input=[chunk.text for chunk in chunks],
)
vectors = [d.embedding for d in resp.data]
Prompt template
text
You are an expert assistant. Answer based ONLY on context.
If the answer is not in the context, say "I don't know".
[Context]
{retrieved_chunks}
[Question]
{user_question}