Senior/Staff Machine Learning Research Scientist Generative AI: Scale AI

At Scale, we believe that the transition from traditional software to AI is one of the most important shifts of our time. Scale is uniquely positioned at the heart of the field of AI as an indispensable provider of training and evaluation data and end-to-end solutions for the ML lifecycle. Our products power the world's most advanced LLMs, generative models, and computer vision models. We are trusted by generative AI companies such as OpenAI, Meta, and Microsoft, government agencies like the U.S. Army and U.S. Air Force, and enterprises including GM and Accenture.

About the Role: Scale’s Generative AI ML team conducts research on models, supervision, and algorithms that advance frontier models for Scale’s applied-ML teams and the broader AI community. In this role, your focus will be on developing new foundational models, algorithms, and forms of supervision for Generative AI. You will be involved end-to-end from the inception and planning of new research agendas to creating high-quality datasets, implementing models, and producing high-caliber publications. You will lead the writing, publishing, and adoption of your work internally with applied teams.

You will:

Publish new methods that advance frontier models/LLMs via human in the loop.

Release papers, datasets, and open source code that improve state of the art open source models.

Evaluate, adapt, and develop new state of the art language and/or multimodal foundation models.

Requirements:

A PhD in AI, Machine Learning, Computer Science, or a related field.

A track record of high-caliber publications in peer-reviewed machine learning venues (e.g. NeurIPS, ICLR, ICML, EMNLP, CVPR, AAAI etc.).

At least 3 to 5 years of model training and evaluation experience.

Strong skills in NLP, LLMs, and deep learning.

Solid background in algorithms, data structures, and object-oriented programming.

Experience working with cloud technology stacks (e.g., AWS or GCP) and developing machine learning models in a cloud environment.

Strong high-level programming skills (e.g., Python), frameworks, and tools such as PyTorch Lightning, Kubeflow, TensorFlow, Transformers, etc.

Interest in capability and alignment research.

Nice to have: Experience dealing with large-scale AI problems (ideally in generative AI) and demonstrated research expertise in post-training methods (instruction tuning, RLHF, tool use, reasoning, agents, etc.).