Government Scheme Awareness

Inspiration We were inspired by the problem that many citizens are unaware of government schemes, legal rights, and public welfare benefits available to them. Information is often difficult to understand, spread across multiple websites, or written in complex language. Our goal was to create an AI-powered platform that makes government awareness simple, accessible, and user-friendly for everyone. What it does Our project is an AI-based Government Awareness Assistant that helps users discover relevant government schemes, understand their legal rights, and access official resources easily. It includes a Scheme Matcher for personalized scheme recommendations, a Legal Advisor for citizen rights awareness, and a Scheme Directory that provides eligibility, benefits, official links, and application steps in a structured format. How we built it We built the frontend using React.js and developed the backend using Flask. The project uses LangChain agents combined with Tavily web search to fetch and summarize information from official sources. We integrated multiple LLM providers including Groq, Gemini, and Cerebras with fallback support to ensure reliability. The complete application was deployed using Render. Challenges we ran into One of the biggest challenges was handling inconsistent outputs from different AI models and preventing raw tool-call syntax from appearing in the frontend responses. We also faced issues while managing multiple model fallbacks, formatting AI responses cleanly, and debugging LangChain agent workflows during deployment. Accomplishments that we're proud of We are proud of successfully building a fully functional multi-agent AI platform focused on social impact. The project can simplify complex government information into understandable responses while using official sources. We also successfully integrated multiple LLM providers and deployed the full-stack application publicly. What we learned Through this project, we learned how AI agents, tool calling, and LangChain workflows operate in real-world applications. We improved our skills in full-stack development, API integration, prompt engineering, deployment, and response handling. We also learned how important reliability and clean output formatting are in public-facing AI systems. What's next for Untitled In the future, we plan to add multilingual support, voice assistance, WhatsApp bot integration, and personalized dashboards. We also want to introduce OCR-based document uploads for automatic scheme recommendations and improve accessibility so the platform can reach a wider audience across India.

Built With

• api
• flask
• html5
• python
• react