About the job
We are now looking for a Senior Research Engineer passionate about Generative AI inference. Are you excited to change the way people infuse AI into products and services? NVIDIA is at the forefront of generative AI models, from language to images. NVIDIA provides building blocks to democratize AI and make generative AI easy to develop, integrate, and deploy. Our team is dedicated to developing optimized inferencing technologies to support our growing generative AI needs. We contribute to all steps of the machine learning lifecycle: from conceptualization, to applied research, engineering for optimized inference, and deployment. Collaborate with research teams, engineers, and open-source community.
Responsibilities
Design and evaluate routing policies for LLM traffic to best use mixture of model systems.
Build and run agentic benchmarks (e.g., Terminal-Bench ) to measure algorithm quality, and turn results into calibration data and routing profiles
Ship to an open-source repo: design docs, code review, docs, and community contributions
Collaborating with engineering teams across all of NVIDIA to ensure our software integrates seamlessly up and down the NVIDIA accelerated serving stack.
Qualifications
Minimum
Bachelor's of Master's degree in Computer Science or equivalent experience.
8+ years of industry experience in Deep Learning frameworks (PyTorch or TensorFlow).
Experience designing or running LLM evaluations/benchmarks — ideally agentic ones — and drawing statistically sound conclusions from them
Understanding of modern techniques in Machine Learning, Deep Neural Networks, Natural Language Processing, or Speech Recognition.
Empirical research mindset: forming hypotheses about new algorithms, running calibrations, iterating on results
Strong communication and interpersonal skills, along with the ability to work in a dynamic and distributed team. A history of mentoring junior engineers and interns is a huge plus.
A desire to constantly grow and learn new things.
Strong computer science fundamentals - algorithms and data structures, computational complexity, parallel and distributed computing, system software.
Preferred
Experience architecting or developing large-scale distributed systems for deep learning.
Agentic benchmark creation and publications.
Knowledge of CPU and/or GPU architecture.
GPU programming (CUDA).