Week 1 (Sept 1–7, 2026) – Problem Validation & Data Collection Interviewed 15 families in Kolhāpur; 12 reported paying at least one incorrect medical bill in the last year. Collected 120 anonymized hospital bill samples from district hospitals and private clinics. Built initial OCR prototype using Google ML Kit; achieved 87% text extraction accuracy on printed bills. Week 2 (Sept 8–14, 2026) – MVP Development Trained custom NLP model on bill dataset to identify line items, codes, and totals (92% accuracy). Launched Flutter app with camera capture, OCR, and English translation. Onboarded first 20 pilot users; average error detection: ₹2,800 per bill. Week 3 (Sept 15–21, 2026) – Multilingual & Error Detection Added Hindi and Marathi translation layers with medical term dictionaries. Implemented anomaly detection for duplicates, upcoding, and phantom charges. Expanded pilot to 50 users; caught ₹3,400 average errors; one user saved ₹8,000 on surgery bill. Week 4 (Sept 22–30, 2026) – Polish & Submission Prep Redesigned UI based on user feedback (larger fonts, simpler navigation for elderly users). Added offline OCR mode for low-connectivity areas. Recorded 2-minute demo video, created Figma mockups, and wrote this Devpost submission. Post-Hackathon Roadmap (Oct 2026 Onward) October: Partner with 3 Kolhāpur hospitals for direct billing API integration. November: Launch insurance claim auto-filing feature. December: Scale to 1,000 users via local NGOs and WhatsApp community groups. Monsoon 2027: Reach 10,000 users across Maharashtra.
Log in or sign up for Devpost to join the conversation.