Senior GPU/HPC Observability & Performance Team Lead — 2 Hours/Day

We are looking for a senior GPU infrastructure, HPC, or ML-platform engineer to technically lead a team of four outstanding junior developers in a selective software-engineering bootcamp. The team is working on a real infrastructure challenge supplied by an established high-tech organization. The company identity, infrastructure environment, and complete operational context will be disclosed only to shortlisted candidates. The technical challenge is to build a compute-efficiency and observability system for GPU workloads. The team will analyze historical and live time-series metrics, identify recurring underutilization patterns, detect issues such as idle accelerators, CPU preprocessing bottlenecks and storage I/O stalls, generate alerts, and produce utilization reports. The project begins with known historical incidents that can be used as ground truth. The goal is to turn those examples into reliable detection logic and then run the detection-alert-report loop continuously. The junior developers will implement the system. You will provide approximately 1-2 hours per working day (Sun - Thu) of hands-on technical leadership.