Job Description
Our client in the Pharmaceutical Manufacturing industry is seeking a Principal Computer Vision Scientist to lead the development of foundation models using biological imaging data, with the goal of accelerating target and biomarker discovery. This team is building a more integrated AI framework for early-stage research, using cellular imaging as a key modality to power a multi-modal foundation model for their in-vitro, high-throughput screening platform.
In this role, you’ll be at the forefront of integrating generative AI into Research & Early Discovery, helping reduce the time from target identification to clinical application.
Day to Day:
Lead the development and deployment of next- AI/ML models using cellular imaging and other biological data types (e.g., molecular, transcriptomics, biomedical literature).
Define the strategy for applying generative AI in early-stage drug discovery, collaborating with cross-functional teams across biology, chemistry, and data science.
Stay current with the latest research in computer vision, deep learning, representation learning, and multi-modal data integration.
Communicate findings through reports, presentations, and scientific publications to both internal and external stakeholders.
Build and maintain collaborations with academic institutions and industry partners.
Must Haves:
PhD in Computer Science, Bioinformatics, Computational Biology, Physics, or a related field.
Professional hands-on experience with pretraining or fine-tuning foundation models for computer vision tasks.
Strong publication record in top-tier ML/CV conferences (e.g., CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML).
Proven expertise in multi-modal data integration and representation learning, ideally applied to biological or pharmaceutical problems.
Advanced programming skills in Python, with SME knowledge of deep learning frameworks in PyTorch, Hugging Face, or PyTorch Lightning.
Proficiency in modern software development practices: version control (Git), continuous integration, testing, and Python packaging (e.g., uv).
Subject matter expertise in foundation models, self-supervised learning, and vision transformers.
Plusses:
Experience with high-content screening, high-throughput data , or single-cell RNA sequencing.
Familiarity with cloud computing platforms (e.g., AWS, Azure, Nvidia DGX Cloud) for large-scale model training and deployment.
Knowledge of systems biology, biophysics, or causal inference in computational biology.
Ability to write well-tested, well-documented code following best practices in machine learning and software engineering.
#J-18808-Ljbffr
