Senior Deep Learning Scientist

Bexorg•New Haven, CT

About The Position

Bexorg is a cutting-edge biotech startup focused on CNS drug discovery, leveraging its proprietary BrainEx platform to conduct preclinical studies on human and porcine brains. We combine advanced wet-lab experiments with AI/ML-driven drug discovery through our XO Digital platform to revolutionize the understanding and treatment of brain disorders. We are seeking a Senior Deep Learning Scientist to design and implement AI models that integrate complex biological signals, including pioneering work in generative graph representation learning and other techniques to model the intricacies of human brain biology.

Requirements

Educational Background: PhD in Computer Science, Machine Learning, or alternative STEM field (e.g., biology or physics) with appropriate experience
Applied Experience: Demonstrated track record of applying deep learning to biological problems (e.g., genomics, transcriptomics, proteomics, or imaging).
Graph & Geometric Deep Learning: Strong practical experience with geometric deep learning and graph neural networks (GNNs); proven ability to tailor these methods to biological data, especially transcriptomics.
Framework Proficiency: Expertise in PyTorch with the ability to build and deploy scalable models.
Multimodal Integration: Experience integrating diverse data types (e.g., transcriptomics, proteomics, mass spectroscopy, etc) using deep learning.
MLOps & Scalability: Familiarity with developing production-quality pipelines, cloud computing, and model deployment best practices.
Collaboration: Excellent communication skills and the ability to work cross-functionally with engineers, biologists and other key stakeholders to convert raw data output into neuroscience discoveries
Industry Exposure: Experience (or strong interest) in drug discovery or biomedical research is a plus.
Research & Innovation: Demonstrated ability to research and implement novel deep learning architectures tailored to complex biological datasets.
Cross-Disciplinary Teamwork: Strong problem-solving skills and ability to work effectively in a cross-disciplinary team (collaborating with bioinformaticians, neuroscientists, experimentalists and engineers).
HPC & Distributed Training: Experience with high-performance computing (HPC) environments or distributed training techniques for large-scale GNN models.
Communication Skills: Excellent communication skills for presenting findings and collaborating effectively with diverse stakeholders.

Nice To Haves

Experience with graph neural networks and generative graph representation learning.
Background or collaborative experience in biological sciences or neuroscience.
Prior experience integrating AI models with high-fidelity biological data.
A publication track record in leading AI/ML conferences or in computational biology/neuroscience journals.

Responsibilities

Model Development: Design, develop, and deploy state-of-the-art deep learning models for analyzing multi-modal biological data.
Integration of Biological Data: Collaborate with bioinformatics, experimental biology and engineering teams to integrate multi-modal datasets into cohesive AI frameworks.
Innovative AI Architectures: Develop deep learning architectures that incorporate biological inductive biases, and explore generative graph representation learning to reveal novel patterns in brain data.
Scalability & Optimization: Optimize deep learning pipelines for petabyte-scale datasets and ensure models are scalable on high-performance computing infrastructures.
Validation & Iteration: Rigorously validate model outputs against biological benchmarks and iterate based on experimental feedback.
Scientific Communication: Publish research findings and present at scientific conferences to contribute to the broader AI and biomedical communities.
Stay Current: Keep abreast of the latest advancements in deep learning and AI, ensuring our models leverage cutting-edge innovations.