Senior Machine Learning Scientist

Role Overview

Senior Machine Learning Scientist focused on model training and maintenance, dataset construction, and model hardening. You will partner with subject matter experts and product owners to define key questions, curate training data, develop/tune models, and communicate insights to peers and stakeholders.

Key Missions

• Model training & maintenance: build, tune, and harden machine learning models for reliable production usage
• Dataset construction: collaborate with experts to define what questions can be answered and create optimal datasets
• R&D on new architectures: contribute to research and development for new model architectures
• Production optimization: improve model performance and operational readiness

Responsibilities

• Develop and tune ML/DL models using best practices for data selection, model architecture, and hyperparameters
• Work cross-functionally with SMEs and product owners to determine key questions and curate data
• Communicate research and model insights to peers and stakeholders
• Optimize models for at-scale, production environments and supporting services

Requirements

• Strong background in machine learning and deep learning, including math/theory
• Experience working with text data for supervised and unsupervised deep learning/ML models
• Language Model familiarity: using, training, and/or fine-tuning, and knowledge of industry LM families
• Software engineering proficiency with at least 8 years experience (Python/SQL/Unix, git/GitHub)
• Experience with experiment tracking and ML tooling (AWS SageMaker, Hugging Face, Transformers, PyTorch, scikit-learn, Jupyter, Weights & Biases)
• DevOps fundamentals (Docker; AWS EC2/Batch/Lambda)
• Strong software development capability for production systems and possible backend API/library development
• Familiarity building lightweight front-ends for demos/POCs (LLMs and/or React/JavaScript/Flask)
• Fluency in written and spoken English
• Education: Master’s degree or PhD in CS/EE/AI/ML/applied math (or outstanding relevant achievements). CS background required (not statistics/pure math) and applied deep-learning orientation.

Nice-to-Haves

• Deep learning experience in other modalities such as vision and speech
• Experience with advanced prompting, fine-tuning, or training LLMs using platforms such as mosaic.ai or stochastic.ai
• Evidence of prior work via website, presentations, open-source code, etc.