생물정보학
pgvector Embeddings — PostgreSQL Vector Search
Use when: writing SQL or Python code for vector similarity search in PostgreSQL, storing/querying embeddings with pgvector, building RAG pipelines over Postgres, creating HNSW or IVFFlat indexes on vector columns, using halfvec/sparsevec/bit types, combining vector search with full-text search (hybrid search), tuning pgvector index parameters (ef_construction, m, lists, probes), choosing distance functions (L2, cosine, inner product), integrating OpenAI/Cohere embeddings with PostgreSQL, or using pgvector with psycopg/SQLAlchemy/Django. Do NOT use when: working with standalone vector databases (Pinecone, Weaviate, Qdrant, Milvus, ChromaDB), using SQLite/MySQL/MongoDB, doing general PostgreSQL administration unrelated to vectors, or working with non-embedding ML tasks like training models or running inference.