Infrastructure & Performance Optimization: Manage and optimize GPU/accelerator infrastructure (on cloud platforms like AWS, Azure, GCP or on-prem NVIDIA clusters) to support high-performance model training and inference. Drive token efficiency, latency reduction, and cost optimization in AI pipelines (through techniques like model quantization, batching, and caching).
Build and iterate AI prototypes quickly by making use of modern AI and automation tools and work closely with each Data Science squad to validate ideas and accelerate delivery.
Synthesize recent work from NeurIPS, ICML, ICLR, and ACL to sharpen active research directions — not just track the field, but translate it into testable ideas
Client & Stakeholder Management: Serve as the primary point of contact for clients; communicate project progress, updates, and outcomes effectively to both technical and non-technical stakeholders.