AI Compiler Engineer

Hyperhunt is hiring an AI Compiler Engineer for one of our clients working at the intersection of AI, ML systems, compiler technology, and specialized hardware acceleration.

They are a deep-tech AI computing company building advanced hardware and software for AI workloads. Founded in 2022, the company is headquartered in Santa Clara, California, with an India presence in Bengaluru

They are looking for engineers with strong experience in compiler development and AI/ML graph optimization who can help optimize and compile machine learning workloads for high-performance AI hardware.

Location: Bengaluru (Hybrid) or Remote

• Strong academic pedigree from premium/top-tier colleges.
• Experience working with top-tier product companies across India.
• Excellent communication and interpersonal skills.

• Design and implement optimizations for AI model execution and graph compilers.
• Improve performance, latency, and hardware utilization for AI workloads.
• Work closely with hardware architects, AI/ML researchers, and software engineers.
• Optimize neural network models using techniques such as: Layer and operator fusion, Graph-level transformations, Quantization, Tiling
• Develop compiler passes and optimizations to transform models from frameworks such as PyTorch and TensorFlow into intermediate representations.
• Work on parsing, semantic analysis, IR generation, and compiler design.
• Contribute to advancements in compiler technology, ML model optimization, and hardware acceleration.
  

• 3-12 years of experience in compiler development, preferably involving AI/ML or graph compilers.
• Strong experience with compiler frameworks such as MLIR and/or Torch-FX.
• Good understanding of: Intermediate Representations (IR), Compiler design, Code generation, Parsing and semantic analysis
• Familiarity with neural network operators and model optimization techniques.
• Understanding of hardware architectures such as GPUs, TPUs, or ASICs.
• Proficiency in C++ and Python or other relevant compiler-development languages.
• Open-source contributions to AI/compiler frameworks would be an added advantage.

If you've worked on the intersection of AI models, compiler infrastructure, and hardware optimization, we'd love to hear from you.