AI SDK Engineer vs AI Semantic Search Engineer

B AI Semantic Search Engineer Only

• Vector embedding model selection, fine-tuning, and evaluation (e.g., text-embedding-3, E5, BGE, GTE)
• Vector database design and optimization (Pinecone, Weaviate, Qdrant, Milvus, pgvector)
• Chunking and document preprocessing strategies for retrieval quality
• Hybrid retrieval combining sparse (BM25) and dense (embedding) search methods
• Retrieval-Augmented Generation (RAG) pipeline architecture and orchestration
• Approximate nearest neighbor (ANN) indexing algorithms (HNSW, IVF, ScaNN)
• Re-ranking and cross-encoder models for precision improvement
• Search quality evaluation metrics (MRR, NDCG, Recall@K, precision@K, end-to-end answer accuracy)