Member of Technical Staff - AI Training Platform

As a Member of Technical Staff, AI Training Platform, you will be a core contributor to the infrastructure that powers our model training ecosystem. You will build and scale the end-to-end tooling required to train, evaluate, and benchmark models. Your work will directly accelerate researchers' ability to map neural networks to novel hardware and push the boundaries of physics-based compute.

• Training Infrastructure: Architect, scale, and maintain the core AI/ML/RL training platform and infrastructure. Design and scale multi-node distributed training systems, implementing elastic sharding and robust data streaming pipelines for fast, large-scale iteration. Implement and robust model checkpointing and recovery mechanisms.
• Framework Development: Maintain and expand our proprietary training framework, ensuring it provides a robust and flexible foundation for all internal model training.
• Optimization & Benchmarking: Develop and optimize kernels using low-level programming models like CUDA and Triton. Design rigorous benchmarking suites to track Model Flops Utilization (MFU), memory bandwidth, and convergence stability.
• Hardware Evaluation Tooling: Build and iterate on tooling to enable rapid, high-level evaluation of novel hardware ideas against benchmarks.
• Cross-Functional Collaboration: Act as a translator, discussing algorithmic trade-offs with theorists and converting model requirements into concrete specifications for infrastructure and hardware engineering teams.
• Optimization & ML Hillclimbing: Design systems to track and visualize quality-vs-efficiency Pareto frontiers, enabling researchers to optimize models for both performance and energy consumption.

• Education: BS in Computer Science, Physics, Electrical Engineering, or Applied Math.
• Experience: 5+ years of experience in AI/ML engineering. Veteran of the modern ML software stack. Demonstrated ability to map state-of-the-art AI model architectures (e.g., transformers, Mixture of Experts, diffusion models) to system performance implication. Deep expertise in how models are partitioned across a cluster, with a mastery of communication primitives, and parallelism strategies.
• Software Development: Strong programming skills in Python or C++. Proven track record of implementing, debugging, and maintaining production-grade training frameworks—such as Megatron-LM, DeepSpeed, Ray, PyTorch Lightning—turning raw compute into a reliable model-building factory.

• MS/PhD or equivalent research/project experience in AI/ML or high-performance computing, with publications
• Experience in training, post-training large-scale LLMs and generative models.
• Experience in one or more of the following technologies: Kubernetes, GPU kernels and performance tuning, LLM inference, CUDA, Triton, NCCL, vLLM, SGLang, VeRL, TRL.

Why Join Us?

• The Mission: Redefine computing for the next 50 years by solving the fundamental energy limitation of AI at a global scale.
• The Impact: Shape the company's future as a foundational team member. Enjoy massive ownership and an outsized opportunity to drive change.
• The Challenge: Dive into deeply complex, intellectually stimulating, and unsolved problems at the cutting edge of multiple, converging fields—you will be defining the future of AI compute.
• The Perks: A comprehensive package including competitive salary, significant equity, best-in-class health benefits, 401k matching, truly unlimited PTO, and complimentary meals when working from our Palo Alto office.

Since 2022, AI has entered the mainstream, reshaping entire industries from education and software development to fundamental consumer behaviors. This revolution has created an unprecedented demand for computation - a demand that is now fundamentally limited by energy, not just in the datacenter, but at a global scale.

At Unconventional, our mission is to solve this. We are rethinking computing from the ground up to build a new foundation for AI that is 1000x more efficient. We're doing this by exploiting the rich physics of semiconductors, mapping neural networks directly to the device physics rather than relying on layers of inefficient abstraction.