Summary

This member of technical staff role focuses on deploying, benchmarking, and optimising large-model inference across heterogeneous accelerators and infrastructure. It is an in-person position based at Callosum's London office, working across orchestration, serving, performance analysis, and production readiness.

Responsibilities

• Run experiments across cloud and on-prem hardware
• Develop reproducible deployment patterns for inference
• Optimise caching, scheduling, batching, and resource allocation
• Integrate accelerator and inference engine upgrades
• Build benchmarking harnesses and regression suites
• Create performance dashboards and prioritise bottlenecks