[Hiring] Applied ML Scientist, Clinical Risk Modelling @Arcanys

Role Description

As an Applied ML Scientist on our GLP-1 companion project, you develop statistical and machine learning models on de-identified electronic health record (EHR) data to predict GLP-1 treatment outcomes. You build, validate, and clinically interpret prognostic models inside our U.S. partner’s secure data environment, working with an obesity-medicine advisor.

In this role you’ll get to:

• Build retrospective patient cohorts from longitudinal EHR data.
• Engineer time-windowed features from medical records, structured to avoid temporal leakage.
• Develop and internally validate statistical and machine learning models for two outcomes: GLP-1 discontinuation and post-discontinuation weight regain.
• Apply clinical-prediction statistics: discrimination, calibration, and subgroup performance, reported to the standards clinicians expect.
• Translate model outputs into interpretable risk tiers with clinical advisors.
• Identify the most promising modelling approaches within a constrained research timeline, balancing scientific rigor with practical impact.
• Document datasets, modelling decisions, and validation results to ensure reproducibility and facilitate future production deployment.
• Collaborate closely with software engineers to ensure models can be integrated into production systems once validated.

Qualifications

• Background in statistics, biostatistics, epidemiology, health data science, machine learning, or a related quantitative discipline.
• Around 3-6+ years developing clinical or tabular predictive models, with strong applied statistics.
• Fluent in tabular ML (gradient boosting, regularized regression) and clinical-prediction methodology, especially calibration and validation.
• Experience engineering features from EHR or other longitudinal records, with a sharp eye for data leakage.
• Fluent in Python and the scientific Python ecosystem, including Pandas, NumPy, scikit-learn, and related libraries.
• Comfortable in restricted clinical data environments and with health-data privacy (HIPAA, GDPR).
• Able to balance scientific rigor with pragmatic decision-making in a fast-paced research setting.

Requirements

• Survival or time-to-event modelling.
• Prior work with EHR or claims data (ICD, RxNorm), ideally in obesity or metabolic disease.
• Experience developing models intended for production deployment.

Benefits

• Top compensation based on expertise – we are only looking for top talent and pay accordingly.
• Long-term work from home + allowances - we let you work from the comfort of your home + amazing allowances you get to enjoy!
• Reimbursable allowance of up to 15K for your home office setup.
• Standard 5 day work week from Monday to Friday.
• Flexible working schedule. You can choose your working hours (within reason).
• Flat management style and open-door policy - no micromanagement and your voice is being heard!
• Work with the latest technologies - you can hone your skills and work with Australian and European clients + a chance to earn more through our self-training program.
• Enrolment to HMO for a maximum of 3 immediate dependents on your 1st day.
• Incentive-based wellness program.
• Toastmasters, English classes, and other learning opportunities.
• Free consultations with Arcanys registered nutritionist-dietitians.
• Free access to our very own Arcanys CrossFit gym (or get a company-sponsored membership in a different gym, whichever suits your needs best).