Teams at Anthropic

Design, build, and operate high-performance distributed inference systems that route requests, autoscale compute across accelerators, deploy new models, integrate new hardware, and tune performance for millions of users while enabling research workflows.

Lead Anthropic's compute expansion in Canada by owning site selection through operations. Coordinate cross-functional teams (leasing, energy, finance, policy, construction), engage government and utilities, manage risks, and translate global compute priorities to Canadian regulatory and market contexts.

Four-month, full-time research fellowship focused on ML systems and performance. Fellows build and evaluate infrastructure-heavy projects (e.g., CPU simulators, accelerator backends, large-scale distributed training), use external models/APIs, receive mentorship, funding for compute, and aim to produce public research outputs.