Principal ML Systems Engineer – AI for Quantum

PsiQuantum • Milpitas, CA

8d

About The Position

The Quantum Architecture Software team at PsiQuantum builds the core software and data systems that underpin the study and simulation of quantum computer architectures and the AI infrastructure and applied systems that accelerate the development of fault-tolerant photonic quantum computers. We sit at the intersection of machine learning, quantum architecture research, and autonomous systems; developing the tools, models, and agent frameworks that let quantum architects move faster and think more clearly.
Our work within the AI for Quantum initiative spans three interrelated areas: training and deploying ML models that operate directly on quantum systems (decoders, scorers, and RL agents for architectural optimization); building scientific agent systems that augment human researchers in exploring large, complex design spaces; and enabling AI-driven control, test, and validation infrastructure for quantum hardware, software, and systems. The team is small, cross-domain, and R&D-oriented. We build things that don't exist yet.

Requirements

• Python, with strong software engineering fundamentals
• PyTorch — training, fine-tuning, and deploying models in production
• Distributed training on GPU clusters (multi-node, NCCL, Slurm or equivalent)
• GPU kernel development — CUDA, ROCm, or Triton; sufficient depth to debug performance and understand memory hierarchy
• AI/ML infrastructure: experiment tracking, model serving, data pipelines
• Experience building or operating autonomous or agentic AI systems

Nice To Haves

• Familiarity with quantum computing or quantum error correction
• Experience in other AI for Science domains (materials, biology, physics simulation)
• Multi-agent system design or orchestration frameworks
• Rust or systems programming experience
• Reinforcement learning, particularly in combinatorial or physical design spaces
• Hardware-adjacent software: FPGA toolchains, embedded control, low-latency inference

Responsibilities

• Develop large-scale pretraining, fine tuning and inference efforts for models used in QEC and related classical control problems.
• Design, implement, and optimize GPU/CUDA/C++ kernels and distributed training pipelines for high-performance model development.
• Collaborate with physicists and engineers to translate quantum phenomena and simulation outputs into ML-relevant data representations.
• Prototype and evaluate new ML architectures for solving QEC-related classical control problems.
• Work closely with infrastructure teams to ensure seamless integration of ML components into existing quantum computation pipelines.
• Drive best practices in ML experimentation, data curation, and reproducible research within the team.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed
• Number of Employees: 251-500 employees