At a Glance
- Tasks: Design and optimise compiler systems for cutting-edge AI technology.
- Company: Join an innovative AI infrastructure start-up revolutionising the industry.
- Benefits: Competitive pay, growth opportunities, and a chance to make a real impact.
- Other info: Be part of a dynamic team driving significant advancements in AI technology.
- Why this job: Tackle complex challenges in AI while contributing to sustainability and efficiency.
- Qualifications: 3+ years in compiler development with expertise in C++ and GPU architecture.
The predicted salary is between 40000 - 50000 £ per year.
- Overview
- 2 days ago Be among the first 25 applicants
This range is provided by Switch Tech Talent. Your actual pay will be based on your skills and experience — talk with your recruiter to learn more.
- Base pay range
- Direct message the job poster from Switch Tech Talent
- Role
- Role: Compiler Engineer (mid, senior, staff)
About the Company
An AI Infrastructure Start-up (B2B Saa S) are building a next-generation platform and seeking Mid/Senior/Staff Compiler Engineers.
Their LLVM/MLIR-based pipeline delivers lower latency, higher throughput, and improved cost efficiency across GPUs, enabling teams to scale production AI without scaling the bill.
Responsibilities
- Designing and extending MLIR dialects, building conversion pipelines.
- Implementing and optimising passes.
- Extending and maintaining backends, curating and writing custom kernels (CUDA/HIP) with runtime integration for optimal occupancy and bandwidth use.
- Profiling, measuring, and preventing regressions using tools such as nsight-compute, rocprof, vtune, and perf, supported by CI’d performance suites.
- Driving upstream improvements to LLVM/MLIR, authoring RFCs, and reviewing PRs.
- Collaborating with model optimisation and runtime teams on ABI, graph partitioning, and kernel launch strategies.
- Experience with LLVM and/or MLIR.
- 3+ years’ experience in production compiler or toolchain development (LLVM/MLIR, GCC, IREE/XLA/TVM, etc.).
- Expertise in C++17/20.
- Strong understanding of GPU architecture.
- Proven performance engineering skills.
- Compiler Engineer Desirable Experience
- Exposure to IREE, Open XLA/XLA, TVM, Tensor RT-LLM, or Glow.
- Knowledge of quantisation (INT8/INT4), dynamic shapes, and autotuning.
- Familiarity with CPU vectorisation (AVX2/AVX-512/SVE), register allocation, or polyhedral techniques.
- Experience with FPGA/HLS flows or custom accelerator backends.
- Why Join
- Opportunity to take ownership of critical compiler subsystems with visible upstream impact.
- Work on one of the most complex and valuable challenges in AI, translating FLOPs into real-world performance.
- Contribute to reducing compute costs, lowering CO₂ emissions, and easing demand on the global power grid.
- Competitive compensation* and growth opportunities within a fast-paced, highly technical team.
- Seniority and Employment Details
- Seniority level: Mid-Senior level
- Employment type: Full-time
- Job function: Information Technology
- Industries: Software Development
- Referrals increase your chances of interviewing at Switch Tech Talent by 2x
Get notified about new Software Engineer jobs in United Kingdom.
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
#J-18808-Ljbffr
Compiler Engineer employer: Switch Tech Talent
At Switch Tech Talent, we pride ourselves on being an exceptional employer that fosters a culture of innovation and collaboration. Located in the vibrant Shoreditch area of London, we offer our Product Engineers not only competitive salaries and benefits but also ample opportunities for professional growth in a dynamic start-up environment. Join us to work on cutting-edge AI solutions while enjoying a flexible work-life balance and a supportive team atmosphere.
