C++ Backend Optimization Engineer
Remote
RemoteFull TimeStartup
Full TimeStartup
Job Summary
Develop highly efficient computation kernels and client-side infrastructure to accelerate neural network inference and DSP pipelines. Design clean, extensible architectures for new framework features while optimizing memory management, concurrency, and performance using techniques like cache miss analysis and thread contention tuning. Leverage understanding of compiler design principles, including ASTs and IR, to refine execution strategies within AI inference engines. Collaborate with an international R&D team to release innovative software for music producers and songwriters.
Required Qualifications
- Strong skills in Modern C++ (C++17/20/23)
- Experience with memory management, concurrency, and performance optimisation techniques
- Understanding of compiler design principles, including Abstract Syntax Trees (AST), intermediate representations (IR) and optimisation passes
- Familiar with computational graph optimization and execution strategies within AI inference engines
- Understanding of basic performance profiling and tuning methodologies (analysing cache misses, memory bandwidth, thread contention)
- Able to design clean, extensible architectures for new framework features
Desired Qualifications
- 3+ years of professional or strong academic/open-source experience related to AI compilers (TVM, MLIR, XLA) or open-source AI inference engines
- Experience with concurrency libraries (oneTBB or OpenMP)
- Experience with GPU programming using Vulkan, Metal, CUDA, SYCL, or OpenCL
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.