Inference System & Performance Engineer

About the Company

Callosum is the Intelligent Systems Company. We believe intelligence emerges from the diversity of co-optimised mechanisms working together. We co-evolve models, workflows, and silicon to build the next era of heterogeneous AI infrastructure.

Responsibilities

• Design and optimise inference serving systems across heterogeneous multi-GPU and multi-node environments
• Own KV cache lifecycle management, batching strategies, and memory allocation to maximise throughput and minimise latency
• Profile and tune GPU kernels, identify bottlenecks across compute, memory, and network, and implement targeted optimisations
• Build and improve scheduling logic for continuous batching, disaggregated prefill/decode, and speculative decoding
• Work with networking primitives like NCCL, NVLink, RDMA, InfiniBand, and RoCE to optimise communication
• Develop tooling for performance visibility, regression detection, and benchmarking across hardware configurations

Requirements

• Deep understanding of LLM inference internals including KV cache, memory management, and attention mechanisms
• Strong systems engineering background with experience optimising distributed GPU workloads
• Proficiency in C++, CUDA, Python, or Rust
• Hands-on debugging skills across GPU, networking, and Linux systems

Preferred Qualifications

• Experience building or optimising production-grade, high-throughput model serving stacks
• Multi-GPU and multi-node inference optimisation using NCCL, NVLink, RDMA, InfiniBand, or RoCE
• GPU memory profiling and CUDA or Triton kernel optimisation
• Linux performance analysis and optimisation

Benefits

• Competitive Salary
• Equity & Ownership
• Private healthcare
• Visa sponsorship and relocation benefits