Machine Learning Healthcare Conference

What we built

Poster Session A:  https://www.mlforhc.org/2025-abstracts

🔸 SlimmAI — Python-based decision-tree model with transparent SHAP explanations
🔸 97 % accuracy identifying cardiovascular-risk patterns in a 100 % Black biomarker cohort
🔸 Maps each risk profile to its ecological tier, so clinicians see where to intervene, not just what to treat.

Why it matters
✔️ Chronic stress + structural racism drive the nation’s highest CVD death gap.
✔️ Conventional risk tools ignore those drivers — leaving Black patients misclassified or unseen.
✔️ SlimmAI closes that blind-spot and gives public-health teams an explainable roadmap for upstream action.

Next steps
• Open-source demo + HRV extension later this year
• Integrating SlimmAI into God’s Little World, our cooperative-economy simulation & research campus breaking ground Dec 2025.

🙏🏾 Grateful to the MLHC reviewers, the Mayo Clinic sponsoring team, and every mentor who believed Black health data could power its own solutions.