-
-
Health Insurance Policy Analysis Made Simple
Inspiration
Buying health insurance is one of the most important financial decisions people make, yet policy documents are often 50–100 pages long, filled with legal jargon, exclusions, waiting periods, and complex clauses. Most consumers rely on agents or marketing brochures and rarely understand what they're actually buying. Post buying when claim is raised - the insurer rejects giving multiple reasons.
I wanted to build an AI assistant that empowers consumers by translating complex policy documents into simple, actionable insights before they make a purchase.
What it does
PolicyLens helps users understand an insurance policy in minutes instead of hours.
Users simply upload a policy brochure or document, and PolicyLens:
Generates a concise AI-powered policy summary. Identifies the key benefits and strengths. Highlights important exclusions, limitations, and risks. Suggests intelligent questions to ask the insurer before purchasing. Explains complex policy clauses in plain English.
Presents everything in an intuitive dashboard that helps users decide whether the policy is suitable for their needs.
Rather than replacing professional advice, PolicyLens enables consumers to make better-informed decisions.
How we built it
I built PolicyLens over week end using CODEX and GPT 5.6 ,OPEN AI PLATFORM as a MOBILE application.
Frontend: Flutter for a responsive cross-platform experience. Backend: FastAPI for document processing and AI orchestration. Document Processing: PDF ingestion and text extraction. Generative AI: Amazon Bedrock to analyze policy content, summarize clauses, identify strengths and risks, answer questions, and explain legal language. REST APIs: Secure communication between the Flutter application and the AI backend.
The solution is designed so that all analysis is generated directly from the uploaded policy document, ensuring users receive insights specific to the policy they are reviewing.
Challenges we ran into
Challenge was balancing AI creativity with factual accuracy. We needed prompts that encouraged the model to remain grounded in the uploaded document and avoid generating unsupported conclusions.
Designing an interface that presents detailed policy analysis without overwhelming users also required careful iteration.
Accomplishments that i am proud of
Built an end-to-end AI-powered health insurance policy analysis mobile application. Successfully transformed lengthy policy documents into concise, understandable insights. Created clause-level explanations that simplify legal terminology. Designed a clean dashboard focused on helping users evaluate policies before purchase. Built a solution that can easily be extended to support multiple insurance providers and policy types.
What i learned
This project reinforced that generative AI delivers the greatest value when combined with domain knowledge and well-structured prompts.
I also learned the importance of grounding AI responses in the uploaded document to improve reliability and user trust.
Finally, i discovered that great user experience isn't about showing more information—it's about surfacing the right information at the right time.
What's next for PolicyLens
My roadmap includes several enhancements:
Extract structured policy information such as insurer, plan name, coverage limits, waiting periods, and optional riders automatically. Get insights from market place , what people saying , regulator are saying ,company news etc Generate a personalized "Should I Buy This Policy?" score backed by explainable AI reasoning. Introduce multilingual support for regional languages.
My vision is to make PolicyLens the trusted AI companion for anyone evaluating health insurance products, enabling consumers to make informed and confident decisions before they buy.
Log in or sign up for Devpost to join the conversation.