Working Student - ML Compiler

Kurz zusammengefasst

Tensordyne is a fast‑growing AI‑system startup that builds high‑performance, low‑power generative‑AI inference platforms. The company designs custom silicon, hardware and software to accelerate multimodal generative AI for hyperscaler and neocloud data‑center customers. With headquarters in Sunnyvale, CA, and Munich, Germany, Tensordyne operates globally and offers comprehensive benefits and a supportive culture focused on people, ethics and ambitious innovation. The opening is for a Working Student in ML Compiler Engineering, based on‑site in Munich, to help create compiler technology for the firm's custom ASICs. In this part‑time role (20 hours per week) the student will spend each day designing and implementing machine‑learning compiler passes and optimizations that match the unique architecture of Tensordyne's ASICs. They will also build intuitive tooling to improve the developer experience of the production ML compiler, and collaborate closely with multidisciplinary groups that include machine‑learning researchers and embedded‑systems engineers. The exact tasks will evolve according to business priorities and the student's own interests, providing a flexible, hands‑on learning environment. The position requires a highly motivated, curious and detail‑oriented individual who is currently enrolled in a Master's program in Electrical Engineering, Robotics, Computer Science or a related AI discipline. Candidates must demonstrate a passion for compilers, a solid grasp of graph theory and graph algorithms, and be proficient in at least one systems programming language such as C++ or Rust. Experience with Python and its ecosystem is also essential. These core qualifications ensure the student can contribute effectively to the full‑stack compiler work and engage with the team's technical challenges. Additional, non‑essential strengths are welcomed. Familiarity with parallel programming frameworks like CUDA or OpenMP, an interest in programming‑language design and compiler theory, and prior experience with performance optimization are all considered valuable assets that can enhance the student's impact and growth within the team. What the role asks for: - Currently enrolled in a Master's program (EE, Robotics, CS or related AI) - Passion for compilers - Strong understanding of graph theory and algorithms - Proficiency in C++ or Rust - Experience with Python ecosystem - Highly motivated, curious and detail‑oriented mindset - Familiarity with CUDA or OpenMP (nice to have) - Interest in programming‑language design and compiler theory (nice to have) - Experience with performance optimization (nice to have)