Role overview
**PreOncology**
is building the world’s first multi-omic, AI-powered platform dedicated to ultra-early cancer detection and prevention. We combine whole-body MRI, liquid biopsy, genomics, lifestyle data, and social determinants of health (SDOH) to define a new specialty in oncology: To prevent, detect and treat Stage Alpha. Our mission is to stop cancer before it becomes clinically detectable.
**Role Overview**
**Location:**
Remote (U.S. only)
**The Role**
Train, tune, calibrate, and evaluate risk-prediction models and implement clean, modular Python code that scales. You’ll work closely with epidemiologists and clinicians, so familiarity with biomedical data (ideally cancer) is essential. This role may be scoped at Junior, Mid, or Senior depending on experience.
**What You’ll Do**
* Own full-cycle ML: data prep/feature engineering → model training → hyperparameter tuning → calibration & reliability → validation & reporting.
* Implement and evaluate performance of survival and risk models.
* Build modular, testable Python libraries and pipelines; write clear interfaces, unit/property tests, and lightweight docs.
* Assess and maintain calibration and drift/recalibration schedules.
* Contribute to Agile rituals; give/receive code reviews.
* Productionize workflows using Nextflow (DSL2).
**Must-Have Qualifications**
* Demonstrated ML depth, not just library usage: can explain why models work, tradeoffs, pitfalls, and how to tune correctly.
* Calibration fluency: how to measure and fix miscalibration; when calibration breaks; implications for clinical thresholds.
* Strong Python (typing, packaging, CI tests, vectorization, profiling) and comfort structuring code for reuse at scale.
* Experience with biological/medical data (preferably oncology).
* Familiarity with Agile collaboration.
* Workflow orchestration with Nextflow (DSL2) for data/ML pipelines.
**How You Work (Culture)**
* Low-ego, collaborative, and kind. You invite feedback and give it constructively.
* Bias-to-action. You thrive in a fast-paced startup, can pivot as concepts evolve, and keep momentum without sacrificing rigor.
* Mission-driven. You can directly relate your past work to our prevention-first mission and articulate that connection clearly.
**Work Authorization**
Applicants must have current, permanent U.S. work authorization. We are unable to sponsor or assume sponsorship now or in the future (including OPT/CPT, H-1B, TN, etc.).
**Compensation**
Competitive salary, benefits, and equity commensurate with experience.
**How to Apply**
Email your resume to Luke.Stetson@preoncology.com with brief answers:
1. A model you tuned that initially under-performed and how you fixed it.
2. How you measured and corrected miscalibration in a past project.
3. Your most relevant biomedical dataset(s) and the pitfalls you navigated.
4. A short example of a Nextflow pipeline you built or extended.
**Why Join Us?**
* Be part of the founding science team shaping a new specialty within oncology.
* Work at the intersection of AI, epidemiology, and clinical translation.
* Collaborate with leaders in machine learning, imaging, and oncology.
* Equity and competitive compensation package.
* Mission-driven: help prevent cancer before it starts.