AI Algorithm Engineer (Virtual Cell)

Quantum Life is advancing AI-driven longevity medicine by developing computational models that help us understand the mechanisms of human aging. CORTEX-Cell is our AI initiative focused on modeling single-cell states and predicting their dynamic evolution over time.

We are seeking an AI Algorithm Engineer with a strong foundation in machine learning and deep learning to develop and optimize our virtual cell models. You will work on model architecture design, pretraining, and temporal prediction, exploring Transformer-based architectures, sequence modeling, and self-supervised learning to build next-generation biological foundation models capable of learning cellular dynamics.

Key Responsibilities

1. AI Model Architecture and Development

   • Design, implement, and optimize deep learning architectures for single-cell representation learning and cellular state prediction.
   • Explore Transformer-based architectures, attention mechanisms, temporal encoders, and other relevant modeling approaches.
   • Develop scalable model architectures for learning complex cellular representations and biological dynamics.

2. Large-scale Pretraining and Self-supervised Learning

   • Develop pretraining strategies using large-scale single-cell transcriptomics datasets.
   • Design training objectives that enable models to learn meaningful cellular representations from heterogeneous biological data.
   • Optimize training pipelines, loss functions, and model configurations to improve learning efficiency and predictive performance.

3. Cellular Dynamics and Aging Trajectory Prediction

   • Develop temporal models that learn the relationship between a cell's current state, elapsed time, and its future state.
   • Investigate cellular aging trajectories and model how cellular states evolve across time, cell types, and biological conditions.
   • Evaluate the model's ability to generalize to unseen donors, datasets, and cellular populations.

4. Model Evaluation and Research

   • Establish rigorous benchmarking and validation pipelines for representation learning and temporal prediction.
   • Analyze model behavior, identify performance bottlenecks, and systematically improve generalization.
   • Collaborate with computational biologists and experimental researchers to translate biological questions into machine learning tasks.

Qualifications

• PhD in Computer Science, Artificial Intelligence, Machine Learning, Computational Science, or a related field.
• Strong knowledge of deep learning fundamentals and practical experience with Transformer-based architectures.
• Proficiency in Python and deep learning frameworks such as PyTorch.
• Experience in model training, experimental design, performance evaluation, and systematic debugging.
• Strong problem-solving skills and the ability to independently investigate complex modeling challenges.

Preferred Qualifications

Experience in foundation models, large-scale pretraining, self-supervised learning, or sequence modeling.

• Experience with single-cell RNA sequencing, gene expression modeling, computational biology, or other biological data.
• Experience with distributed training, multi-GPU systems, or large-scale deep learning infrastructure.
• Research publications or contributions to relevant open-source projects.

What We Offer

• The opportunity to develop AI models for fundamental biological problems rather than conventional prediction tasks.
• Ownership of core model architecture and training methodology in an emerging virtual cell research program.
• A highly interdisciplinary environment combining AI, computational biology, and longevity research.
• The opportunity to contribute to research with the potential for scientific publication and long-term technological impact.