Senior / Staff Software Engineer (AI / Compiler)

JR United Kingdom; Bedford; On-site; GBP 145,000 - 180,000; Full time

Job summary

A leading company in AI accelerators is seeking experienced Senior and Staff Software Engineers to develop high-performance computing infrastructure. The ideal candidates will design systems for low-latency inference and work closely with both hardware and ML teams to enhance AI platform performance. This position offers a competitive salary starting from £145k+, stock options, and opportunities to shape the AI of the future.

Benefits

Stock options

Competitive salary

Qualifications

• 5+ years in performance-critical systems (HPC, HFT, AI).
• Experience with ML compilers and runtime optimizations.
• Strong skills in debugging and performance tuning.

Responsibilities

• Design and build high-performance systems for AI/ML workloads.
• Optimize systems for ultra-low latency and real-time inference.
• Collaborate with hardware teams to leverage OTPU capabilities.

Skills

C++

Python

Debugging

Profiling

Performance Tuning

Distributed Systems

Real-time Data Processing

Education

Degree in Computer Science

Degree in Engineering

Degree in Mathematics

Tools

LLVM

MLIR

PyTorch

ONNX

OpenXLA

Job description

Social network you want to login/join with:

Senior / Staff Software Engineer (AI / Compiler), Bedford

Company Overview

Flux is pioneering a new class of AI accelerators called Optical Tensor Processing Units (OTPUs). We’ve already developed functioning prototypes and are now scaling our operations in London. Our work environment rewards innovation, speed, and bold thinking.

The role

We’re hiring Senior and Staff Software Engineers to build the high-performance computing infrastructure that powers our OTPUs. This role involves rethinking how AI workloads are executed at speed and scale, focusing on low-latency, high-performance systems.

You’ll lead the design and implementation of software systems for distributed, low-latency inference across clusters. Collaborate closely with hardware and ML teams to optimise model representation, execution, data movement, and scheduling. Your work may involve compiler techniques, systems tuning, or custom runtime design, shaping the performance of our AI platform. Ideal candidates think in microseconds and have experience in HFT, large-scale scientific compute, or AI infrastructure.

Responsibilities

• Design and build high-performance systems for AI/ML workloads across distributed clusters
• Optimize for ultra-low latency and real-time inference—profiling, tuning, rewriting systems as needed
• Identify and resolve performance bottlenecks across the stack
• Collaborate with compiler engineers to improve code generation and execution paths using tools like LLVM or MLIR
• Work with hardware teams to leverage OTPU architecture capabilities
• Extend ML frameworks (e.g., PyTorch, ONNX, OpenXLA) for performance-critical inference
• Lead design reviews, mentor engineers, and promote HPC best practices
• Stay updated on AI infrastructure, compute systems, and compiler tooling developments

Skills & Experience

• 5+ years in performance-critical systems in HPC, HFT, simulation, or AI infrastructure
• Deep understanding of distributed systems, especially real-time data processing
• Strong skills in C++ and Python for performance applications
• Experience with ML compilers like LLVM, MLIR, and runtime optimizations
• Knowledge of ML frameworks like PyTorch, ONNX, OpenXLA, and their optimization
• Experience scaling AI workloads on clusters or custom infrastructure
• Excellent debugging, profiling, and performance tuning skills
• Degree in Computer Science, Engineering, Mathematics, or related field

Details

• Competitive salary from £145k+, depending on experience
• Stock options in a growing AI company
• Based in our new 5,000 sq. ft. office in Kings Cross, London
• Candidates within a 45-minute commute preferred; an extra £24k/year incentive offered for living within 20 minutes