HealthLeap is an AI start-up revolutionizing healthcare through predictive analytics, initially focused on disease-related malnutrition—a critical form of patient deterioration affecting virtually every hospital condition. Our mission is to maximize health outcomes globally by building a scalable AI platform that screens patients comprehensively using electronic health records (EHRs), labs, clinical notes, and more.
Requirements
- Statistics: parametric and non-parametric tests, hypothesis testing, experimental design, confidence intervals, and causal inference basics.
- ML fluency: Python, SQL; polars (or pandas), scikit-learn, XGBoost/LightGBM (PyTorch/transformers a plus); survival/time-to-event experience is great.
- Visualization & storytelling: Expert at turning complex analyses into crisp user visualizations, dashboards, and narratives for clinicians and executives.
- Read the latest research and rapidly translate new statistical/ML papers into pragmatic wins.
- Understanding of fairness: Independence, Separation, and Sufficiency
- Uncertainty quantification
- Covariate and prediction drift detection in production
Responsibilities
- Own end-to-end modeling from financial incentives and problem framing to a validated model.
- Estimate impact with rigorous retrospective analyses (LOS, readmissions, mortality, reimbursement).
- Productionize pipelines and rollouts with reliability.
- Monitor & improve: drift, calibration/uncertainty, and fairness (Independence/Separation/Sufficiency).
- Translate research into pragmatic wins for our platform.
- Partner with stakeholders: clear visuals, crisp narratives, and method presentation for analysts, clinicians, and executives.
Other
- Passionate about AI's potential in healthcare; outcomes-oriented with a focus on impact, not just research.
- Customer-facing: Comfortable interviewing stakeholders, presenting to AI/data science leaders, and defending methods.
- 3 - 5+ years of relevant experience from a high-growth environment.
- Resourceful, fast learner, high ownership, bias to action, fast experimentation cycles, and ability to work independently while collaborating in a small team.
- Background in applied AI companies with strong product traction (not hype-driven firms).
