Leads development of performance optimizations for ML models across graph, kernel, and system levels, advances Thunder compiler with new passes and tools, and ensures seamless PyTorch Lightning integration. Requires strong PyTorch expertise, optimization techniques, and distributed systems knowledge.
Develop performance-oriented model optimizations at multiple levels:
Advance the Thunder compiler by building optimization passes, graph transformations, and integration hooks to accelerate training and inference workloads.
Work across the software stack to ensure optimizations are accessible to end users through clean APIs, automated tooling, and seamless integration with PyTorch Lightning.
Design and implement profiling and debugging tools to analyze model execution, identify bottlenecks, and guide optimization strategies.
Collaborate with hardware vendors and ecosystem partners to ensure Thunder runs efficiently across diverse backends (NVIDIA, AMD, TPU, specialized accelerators).
Contribute to open-source projects by developing new features, improving documentation, and supporting community adoption.
Engage with researchers and engineers in the community, providing guidance on performance tuning and advocating for Thunder as the go-to optimization layer in ML workflows.
Work cross-functionally with Lightning’s product and engineering teams to ensure compiler and optimization improvements align with the broader product vision.
Anticipated annual base salary range: $225,000—$275,000 USD
In addition to base salary, total rewards package includes discretionary bonus, equity component, and comprehensive benefits.
Conducts ML research on LLMs and audio models to enhance real-time voice agents' reasoning, latency, and conversational quality. Prototypes models, designs evaluations, and bridges research to production systems requiring strong PyTorch expertise and experimental mindset.
Owns end-to-end ML research initiatives developing novel architectures, training methods, and evaluation for clinical intelligence using longitudinal patient data. Requires strong ML foundation, PyTorch experience, and ability to drive ambiguous high-stakes problems to validated results.
Owns end-to-end production ML systems for clinical workflows, including training/fine-tuning LLMs for medical reasoning and question answering. Requires strong ML/software engineering, PyTorch experience, and ability to handle high-stakes ambiguity with real patient impact.
Research Engineer advancing video/image generation models for AI characters, leading fine-tuning, novel architectures, data pipelines, and optimizations using PyTorch and multimodal techniques. Requires expertise in generation models and distributed training.
Develops alignment algorithms, data pipelines, and sampling methods to optimize post-training AI models for performance and efficiency. Requires PhD or equivalent, ML expertise including reinforcement learning and transformers, and production code experience.