ClaimShield AI — Project Overview:
ClaimShield AI is an AI-powered health insurance policy audit and comparison platform. Instead of acting like a generic chatbot or simple text summarizer, it performs an Adversarial Claim Audit—thinking like an insurance auditor to identify hidden fine print, financial traps, strict deadlines, and missing evidence before a claim gets delayed or rejected.

Core Features & Workflows:
Adversarial Claim Audit (/ & /report)

Upload & Intake: Users upload their policy PDF and describe their hospital admission or claim incident.
Evidence-Backed Analysis: Identifies hidden risks like room-rent caps, proportionate deductions, mandatory 24h notification deadlines, co-pay requirements, and exclusions.
Actionable Output: Generates live deadline countdown timers, a document collection checklist, and an AI-assisted claim notification / dispute email draft (“Magic Email”).
Multi-Policy Comparison (/compare)

Side-by-Side Audit: Upload 2 to 3 health insurance policies (PDFs) to compare critical claim conditions side-by-side.
Condition Matrix: Compares Sum Insured, Room Rent Limits, Co-Pay %, PED Waiting Periods, Claim Deadlines, ICU Caps, and Ambulance Coverage.
Priority-Tailored Insights: Allows users to select personal priorities (Lower out-of-pocket cost, Shorter waiting periods, etc.) to highlight key differences with grounded policy clause citations.
Trustworthy & Objective: Uses evidence-backed indicators with zero fake scores or misleading percentage win probabilities.
Technology Stack:
Frontend: React (Vite), Vanilla CSS (Sleek, dark-mode healthcare & financial security design system).
Backend API: Python FastAPI (RESTful API & ephemeral file parsing).
Document AI: Gemma (via Gemini API) for structured clause extraction & document reasoning.