Senior Software Engineer - Kernels (AI Hardware)

Location: Hybrid, onsite at the Belgrade, Serbia office 3–5 days per week

Company

hirify.global develops software and hardware solutions for generative AI and AI compute engines.

What you will do

• Develop, enhance, and maintain software kernels for next-generation AI hardware.
• Productize the software stack for an AI compute engine.
• Map algorithms and computational graphs from AI frameworks to underlying hardware architectures.
• Optimize hardware-software co-design trade-offs across the full-stack toolchain.
• Collaborate with compiler, ML, systems, mixed-signal, DSP, and CPU experts to build compiler infrastructure.
• Scale software deliverables within tight development windows.

Requirements

• MS in computer engineering, mathematics, physics, or a related field with 5+ years of industry experience, or a PhD in a related field with 1+ years of industry experience.
• Strong understanding of computer architecture, data structures, system software, and machine learning fundamentals.
• Proficiency in C/C++ and Python development in Linux environments using standard development tools.
• Experience implementing algorithms for specialized hardware such as FPGAs, DSPs, GPUs, and AI accelerators, including CUDA-based development.
• Experience implementing ML operators such as GEMMs, convolutions, BLAS, softmax, layer normalization, pooling, and SIMD operations.
• Experience developing for embedded SIMD vector processors such as Tensilica, plus ownership and leadership skills.

Nice to have

• Startup, small-team, or incubation experience.
• Experience with TensorFlow or PyTorch and ML models for computer vision, NLP, or recommendation systems.
• Experience with ML compilers and related technologies such as MLIR, LLVM, TVM, or Glow.
• Experience at a cloud provider or AI compute/subsystem company.

Culture & Benefits

• Respectful, collaborative, and inclusive working environment.
• Direct communication and emphasis on humility.
• Opportunity to work across software, compiler, ML, systems, and hardware disciplines.
• Focus on execution, ownership, and solving challenging AI compute problems.