• Architect and maintain client-specific and internal RAG pipelines, including embedding generation, document chunking, and metadata tagging • Collaborate directly with enterprise clients to assess data landscapes, retrieval requirements, and grounding needs • Select, test, and optimize embedding models (text, multimodal) for accuracy and efficiency • Design retrieval strategies (dense, sparse, hybrid) to maximize precision and recall • Establish and monitor evaluation metrics for retrieval performance (e.g., recall@k, MRR, nDCG) • Align retrieval structure, granularity, and metadata with reasoning workflows • Maintain documentation, schemas, and retrieval guidelines for internal and client teams • Stay current with advancements in vector databases, embedding models, and retrieval frameworks