Machine Learning Scientist, Large Language Models

Flagship Pioneering, Inc.•Cambridge, MA

66d•$136,000 - $240,000

About The Position

Pioneering Intelligence (PI) is a Flagship initiative focused on developing and harnessing AI/ML based platforms for innovative applications in life sciences and beyond.â¯We are seeking a creative and talented machine learning scientist to join us and create novel LLM-powered systems aimed at accelerating Flagship's scientific innovation and platform creation. You will be joining a team with diverse expertise building state-of-the-art LLM systems that aim to accelerate innovation across Flagship's entire venture system.

Requirements

PhD in a relevant field such as machine learning, computer science, math, statistics, physics, or a closely related quantitative field
Demonstrated research excellence in ML for language models or adjacent areas (reasoning/agents, sequence modeling, representation learning, optimization), evidenced by publications at top venues or other impactful research work
Experience with modern machine learning frameworks, e.g. PyTorch or JAX, including reproducible experiment workflows
Fluency in Python and standard ML tools and packages
Ability to take initiative, work independently, and collaborate effectively within a team

Nice To Haves

Hands-on experience building and evaluating LLM systems, including planning/tool-use frameworks, multi-agent simulations, memory & reflection loops, retrieval-augmented generation/tool integration, instruction tuning, preference optimization (RLHF/RLAIF/DPO) and evaluation methodologies (LLM-as-judge, human studies)
Experience and knowledge in an applied science domain (e.g. chemistry, biology, physics, materials, etc.)
1+ year industry experience
Experience with modern software development and cloud services, e.g. Git, Docker, AWS Batch, Step Functions, EKS, SageMaker, etc.

Responsibilities

Design and prototype novel LLM-based workflows that accelerate scientific innovation across diverse domains
Benchmark and evaluate prototypes, defining success metrics and designing custom benchmarks tailored to each domain
Work with our ML Engineers to scale promising prototypes
Track and synthesize state-of-the-art ML research and validate findings with careful experimental design and statistical analysis
Supervise development of production-quality feedback loops for user testing and feedback