Strong experience with deep learning frameworks (PyTorch, TensorFlow) for production ML systems, including designing and deploying ranking models, personalization engines, and real-time inference pipelines with knowledge of model serving patterns (e.g., TorchServe, TensorFlow Serving, BentoML) and online/offline evaluation frameworks. Experience building and deploying neural search and embedding-based retrieval systems, including vector databases (e.g., Pinecone, Weaviate, FAISS), semantic search architectures, and large-scale similarity search at scale with sub-100ms latency requirements.