Senior / Staff Machine Learning Research Engineer

Calico•South San Francisco, CA

8d•$220,000 - $290,000•Onsite

About The Position

Calico seeks Machine Learning Research Engineers to join our rapidly growing ML team. You will play a critical role establishing the engineering culture for frontier ML research in drug discovery. You will bring a high level of engineering rigor to our machine learning efforts, and accelerate our research maturing into tangible clinical and product impact. This will be a high agency role designed for a builder who wants to operate as a founding member of a new functional group. No biology or life sciences background is required for this role.

Requirements

A strong intellectual curiosity for life sciences
BS/MS with 7+ years or PhD with 4+ years of relevant ML software engineering experience in industry or academia
Expertise in Python and JAX or PyTorch
Hands-on experience building, training, or optimizing advanced ML architectures (e.g., transformers, diffusion networks, GNNs)
Experience driving complex machine learning engineering projects from concept to production
Must be willing to work onsite at least four days a week

Nice To Haves

Advanced degree in computer science or a relevant field
Experience working with biological, medical, or chemistry datasets
Contributions to open-source ML projects or relevant academic publications
Experience training models across distributed systems and optimizing performance
Experience with cloud infrastructure (e.g., GCP, Kubernetes, Docker)

Responsibilities

Drive the engineering vision behind our machine learning models, from identifying high-impact research engineering opportunities to delivering production-grade systems.
Proactively identify emerging engineering gaps required for expanding our research capabilities.
Architect solutions for complex systems challenges, such as asynchronous execution, hardware orchestration, and high-throughput data pipelines.
Build foundational libraries that enforce software engineering rigor.
Translate prototype models (e.g., diffusion networks, vision transformers, and DNA sequence models) into highly-optimized JAX or PyTorch code on accelerators.
Dive deep into model architectures to optimize training and inference performance, implementing advanced strategies such as model parallelism.
Design and implement evaluation frameworks, benchmarks, and scientific workflow tools that accelerate the research lifecycle and allow the team to rapidly test promising ideas.