Approach

In building EcoMedAI, the team embraced a user-centered, data-driven design philosophy with a focus on real-world impact in healthcare. The goal was to empower hospitals to make sustainable choices in procurement and waste management, by leveraging the latest advances in artificial intelligence and cloud technologies. The project was developed as part of Philly {Codefest} 25, channeling the urgency of environmental sustainability in healthcare into an actionable, intelligent platform.

Vision and Innovation

EcoMedAI envisions a future where hospitals not only deliver outstanding patient care but also lead in environmental stewardship. As the US healthcare sector ranks among the highest in global greenhouse gas emissions, EcoMedAI aims to transform the sector into an exemplar of sustainable practice. The core innovation lies in automating the analysis of hospital supply chains and waste sorting—a task that is prohibitively time-consuming for already overextended staff—making sustainability actionable and scalable.

Identifying Unique Challenges

The project identified several unique challenges:

• Data Complexity: Hospitals use thousands of different supplies, and staff lack the time and tools to evaluate each item's sustainability.

• Vendor Engagement: Manufacturers need incentives to shift toward greener products and packaging.

• Waste Sorting: Missorting medical waste dramatically increases both cost and environmental impact, yet risk aversion and lack of clarity lead to overuse of costly hazardous waste streams.

• User Adoption: The system needed to be easy enough for busy staff to use without adding to their workload, and flexible enough to audit multiple departments.

Resolving Complex Problems

To address these, EcoMedAI implemented:

• Sustainable Supply Recommender: Ingests hospital Bill of Materials (BOM) and recommends more sustainable, affordable alternatives. It uses a combination of FAISS vector stores (for similarity search) and large language models (for intelligent product matching and ranking).

• Waste Classifier: Deploys a FastAPI backend with a ResNet50-based image classifier to help staff correctly sort medical waste according to WHO standards, reducing missorting and costs.

• Automated Audits: The system can rapidly assess waste sorting practices across departments, providing ratings and actionable recommendations.

User-Centric Design

EcoMedAI's design revolves around real hospital workflows:

• Simple Onboarding: Staff can upload supply lists via CSV or through the API, receiving instant, actionable reports.

• Visual Dashboards: The web interface offers clear summaries of cost savings and carbon reductions, with department-level breakdowns.

• Interactive Waste Sorting Coach: Staff can upload images of waste and receive immediate, accurate classification and compliance guidance.

Detailed Pages and Features

• About EcoMedAI: Explains the mission to green healthcare, the critical need for sustainability, and the project’s Codefest origins.

• Sustainable Coach: Guides staff in sorting waste, highlighting both environmental impact and regulatory compliance.

• Procurement Recommendations: Offers actionable, ranked suggestions for sustainable supply purchasing, complete with carbon footprint metrics.

• API and CLI Modes: The backend supports both interactive web/API use and command-line batch processing for integration into existing hospital workflows.

Conclusion

EcoMedAI represents a fusion of cutting-edge AI and practical hospital needs, tackling one of healthcare’s most pressing challenges: reducing its environmental footprint. By automating sustainable procurement and waste management, and providing staff with simple, actionable tools, EcoMedAI creates a multiplier effect—empowering not just a single hospital, but driving broader change in the industry. The project’s commitment to open-source principles, accessibility, and user-centered design ensures that its impact will continue to grow, inspiring sustainable practices in healthcare for years to come.