Machine Learning Scientist

About Us

Our mission is to cure cancer through high performance, accessible early cancer detection. That means saving lives.

Delfi Diagnostics is a Johns Hopkins spinoff focused on the non-invasive detection of cancer at earlier stages, when it is most curable. DELFI uses artificial intelligence and whole-genome sequencing to detect unique patterns of DNA fragmentation in the blood of patients with cancer. These analyses are performed through simultaneous examination of millions of DNA sequences using machine learning to identify tumor-specific abnormalities.

In our passionate pursuit to radically improve health outcomes, we serve humanity when we:

Lead with Science, Anchor in Pragmatism: We pioneer life-changing science by ensuring quality, transparency, and rigor at all times.

Build With & For All: We embrace diverse backgrounds to innovate and achieve together. We are not just building a product—we aim to disrupt the path of cancer for all, no matter geography or socioeconomic class.

Put We over I: We are a home for high-performing people. Through teamwork, we build collective intelligence. Each of us wins when those we serve and those who serve with us win. We show up with empathy, humility, and integrity at every step of the journey.

About the Role

In this role, you will develop, tune, and advance Delfi’s machine learning models for early cancer detection. You’ll focus on improving model performance through structured experimentation, creative modeling strategies, and rigorous benchmarking—pushing models up internal leaderboards and identifying what drives improvement.

You’ll also explore how the raw intelligence of large language models (LLMs) can be applied to improve model performance and feature representations—leveraging their reasoning capabilities. Working closely with bioinformaticians, engineers, and data scientists, you’ll operate at the intersection of machine learning and biology, translating genomic signals into clinically meaningful insight.

This role is ideal for a scientist who enjoys hands-on modeling, thrives on iteration and discovery, and seeks to combine deep technical understanding with curiosity about new forms of machine intelligence. While prior industry experience is preferred, we also welcome exceptional PhD graduates or postdocs who have demonstrated strong applied ML engineering skills and a track record of collaborative, reproducible work.

• You’ll design, implement, and optimize machine learning models for genomic and fragmentomic data, perform systematic benchmarking to assess model quality, and analyze the factors that drive predictive improvements.
• You’ll explore the use of LLMs to enhance feature representations and model architectures.
• You’ll ensure robust, reproducible experimentation through sound data practices and MLOps best practices such as versioning, model tracking, and environment management.
• You’ll collaborate closely with teams across computational biology, bioinformatics, and software engineering to build shared understanding and integrate insights from data.

• You will have led multiple modeling efforts that improved performance on Delfi’s internal leaderboards through careful experimentation and analysis.
• You will have evaluated and demonstrated how LLMs can strengthen feature representations or model architectures to achieve measurable performance gains.
• You will be recognized as a rigorous and creative contributor—someone who combines scientific depth with curiosity and collaboration to drive the frontier of ML for early cancer detection.

• MS or PhD in Computer Science, Machine Learning, Computational Biology, Applied Mathematics, or a related field
• Experience developing and evaluating ML models in applied or collaborative research settings, with a demonstrated ability to deliver high-quality, maintainable code and reproducible results
• Experience working in team-based environments with shared codebases and version control practices
• Proficiency in Python, including use of ML frameworks such as PyTorch, TensorFlow, or scikit-learn
• Experience applying MLOps best practices for experiment tracking, model versioning, or data pipeline reproducibility (e.g., MLflow, Weights & Biases, or equivalent)
• Demonstrated success improving model performance through experimentation, architecture design, or advanced optimization methods
• Familiarity with large language models (LLMs), including APIs, frameworks, and fine-tuning methods
• Strong grounding in statistics, data analysis, and reproducible experimentation
• Excellent communication skills and the ability to collaborate effectively across scientific and technical disciplines

• Experience with genomic, sequencing, or other biological data
• Exposure to cloud-based ML environments (AWS, GCP) or large-scale data pipelines
• Background in deep learning, probabilistic modeling, or ensemble methods
• Record of research publication, technical presentations, or open-source contributions

An Equal Opportunity Employer

We are an equal opportunity employer and value diversity at our company. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.