AI-Powered Platform for Streamlined Access to Government Schemes (SchmeSaathi)

---

Video Link : https://drive.google.com/drive/folders/13FLs1phKQ8f9wYZBmXN0eV-1LfwwfNSl?usp=drive_link

Repository Link : https://github.com/indu-shekhar/SchemeSaathi

🛠️ Problem Identified

Key Challenges:

1. 🎯 Scattered Information

   Citizens struggle to find relevant scheme details as they are spread across various platforms.

2. ⚙️ Complex Eligibility Requirements

   Understanding eligibility criteria and required documents can be daunting.

3. 🌍 Language Barriers

   Limited multilingual support prevents rural and underserved communities from accessing schemes.

4. 🚧 Cumbersome Processes

   Inefficient application workflows make it difficult to secure benefits.

These challenges result in many citizens missing out on critical opportunities to improve their lives.

---

🚀 Methodology

💡 Value Proposition

Our platform tackles these challenges by:

• Simplifying Access: Consolidating information on over 2,500 schemes for easier discovery.
• Personalized Recommendations: Using AI-driven matching based on user profiles.
• Enhanced Accessibility: Providing multilingual and voice-enabled chatbot support for inclusivity.

---

📌 Our Approach

1️⃣ User-Friendly Interface

• Secure login with phone numbers and OTP verification.

2️⃣ Document Processing

• OCR technology extracts data from documents for automated profile creation.

3️⃣ Scheme Matching

• Advanced SentenceTransformer embeddings and ChromaDB vector search.

4️⃣ Dynamic Updates

• Real-time synchronization of scheme information.

5️⃣ Chatbot Assistance

• Multilingual, voice-enabled chatbot for personalized guidance.

---

🔧 Technologies Used

Technology	Details
Backend	Flask, Python
AI Components	SentenceTransformer, ChromaDB, Meta-Llama-3-8B-Instruct
Frontend	HTML, JavaScript, Tailwind CSS
Conversational AI	LangChain, Hugging Face
Additional Tools	OCR for document extraction

---

🌈 Addressal of the Problem

Our platform solves these issues by:

1. Scattered Information: Offering a centralized data repository.
2. Complex Eligibility: Using AI to analyze user profiles and suggest schemes dynamically.
3. Language Barriers: Empowering non-native speakers with multilingual, voice-enabled chatbots.
4. Cumbersome Processes: Simplifying workflows for smooth user experiences.

📊 Impact Metrics

• ⏱️ Time-Saving: Faster identification of relevant schemes.
• 🌐 Accessibility: Increased inclusivity for diverse language users.

---

💰 Financial Outlay

Component	Estimated Cost (INR)
AI Model Development	15,000
Backend & Frontend Development	10,000
Database Integration	5,000
Multilingual Chatbot**	35,000
Infrastructure (Servers, APIs)	15,000
Marketing & Outreach**	30,000
Contingency	2,000
Total	47,000

---

Timeline

Phase	Duration	Activities
Phase 1: Planning	1 month	Requirement analysis, stakeholder consultation, and project roadmap creation.
Phase 2: Development	3 months	Backend, frontend, AI model development, and database integration.
Phase 3: Testing	2 months	Functional, user acceptance, and performance testing.
Phase 4: Deployment	1 month	Platform launch and initial user feedback collection.
Phase 5: Optimization	Continuous	Feedback-based refinements and feature upgrades.

Members:

Indu Shekhar Jha (122CS0076)

Ananya Srivastav (122CS0077)

Abhishek kumar (122AD0039)