Senior AI Engineer with 10 years building production ML, retrieval, and distributed inference systems across Cohere, Google, Labelbox, and PathAI. Specialized in enterprise retrieval, agentic AI workflows, multimodal document understanding, and distributed ML infrastructure operating under real-world constraints where grounding, latency, traceability, and reliability directly affect production outcomes. At Cohere, built hybrid retrieval and multi-step reasoning systems using embeddings, reranking, structured tool routing, and evaluation pipelines for enterprise AI workflows. At Labelbox and PathAI, developed multimodal data and inference systems handling noisy OCR, heterogeneous clinical datasets, pathology imaging, and production-scale document processing. Earlier at Google, built distributed ML infrastructure, backend services, and large-scale data processing systems across Azure cloud environments. Experienced owning systems end-to-end from ingestion and retrieval pipelines through production inference behavior, observability, evaluation, and reliability improvements
