If the registration will be completed over a phone call by an AI voice agent, the design should be conversational rather than form-based. The AI should collect information naturally, verify the visitor’s identity using Aadhaar and mobile number, and then create the registration in the backend exactly as the website does. ⸻ AI Voice Agent Requirement Document Kumbh Mela 2027 Visitor Registration Objective Develop an AI Voice Agent that can register visitors for Kumbh Mela 2027 over a phone call. The AI should: * Welcome the visitor. * Explain the purpose of the call. * Collect all required information naturally. * Validate Aadhaar number and mobile number. * Confirm details before submission. * Generate the Kumbh Pass. * Inform the visitor about successful registration. * Trigger SMS/WhatsApp confirmation. The AI should behave like a polite government helpdesk executive, not a telemarketer. ⸻ Registration Flow Incoming Call │ Greeting │ Consent to Register │ Collect Aadhaar Number │ OTP Verification │ Fetch Aadhaar Details (if integrated) │ Confirm Name │ Collect Remaining Details │ Repeat Information │ Confirmation │ Create Registration │ Generate Pass Number │ Generate QR Code │ Send SMS / WhatsApp │ End Call ⸻ AI Conversation Script 1. Greeting Namaste. Welcome to the official Kumbh Mela 2027 Visitor Registration Helpline. I am your virtual registration assistant. I will help you register for your visit. This process will take about three to four minutes. May I continue? Possible responses: * Yes * Sure * Continue If No: Thank you. Please call again whenever you are ready. Have a wonderful day. ⸻ 2. Aadhaar Collection Please tell me your 12-digit Aadhaar number. Example: 1234 5678 9123 Validation: * Must be 12 digits. * Numeric only. * Verify checksum (Verhoeff algorithm). * Reject invalid format. If invalid: The Aadhaar number appears to be incorrect. Could you please repeat it? ⸻ 3. OTP Verification After receiving Aadhaar: Backend sends OTP. AI says: I have sent a six-digit verification code to your registered mobile number. Please tell me the OTP. Validate OTP. If incorrect: The OTP does not match. Would you like me to send a new OTP? Maximum attempts: 3 ⸻ 4. Aadhaar Data Fetch (Optional) If UIDAI/eKYC integration exists: Automatically fetch: * Name * Gender * Date of Birth AI: I found your Aadhaar details. Your name is Rajesh Kumar. Is this correct? If Yes Continue. If No Transfer to human support. ⸻ 5. Mobile Number Confirmation If Aadhaar mobile available: Is this mobile number your current contact number? Otherwise: Please tell me your mobile number. Validate: * 10 digits * Indian mobile format ⸻ 6. Emergency Contact Please provide an emergency contact number. ⸻ 7. State Which state are you travelling from? Example: * Uttar Pradesh * Bihar * Maharashtra Or Country if international. ⸻ 8. Arrival Date On which date are you planning to arrive at Kumbh Mela? Validation: Future date only. ⸻ 9. Mode of Transport How will you travel? Choices: * Train * Flight * Bus * Personal Vehicle ⸻ 10. Stay Type Where will you stay? Options: * Hotel * Tent City * Akhada * Day Visit ⸻ 11. Confirmation AI summarizes: Thank you. Please confirm your registration details. Name: Rajesh Kumar Mobile: 98XXXXXXXX Arrival Date: 18 January 2027 Travel: Train Stay: Tent City Is everything correct? Options: * Yes * No If No Edit requested field. ⸻ 12. Registration Backend: INSERT Visitor Generate KM2027-000001 Generate QR Code. ⸻ 13. Success Message Congratulations. Your registration has been completed successfully. Your Kumbh Pass Number is KM2027-000001. A QR Code and registration details have been sent to your mobile number through SMS and WhatsApp. Please carry your Aadhaar card and this pass while travelling. We look forward to welcoming you to Kumbh Mela. Thank you. ⸻ Database Fields Field Source Pass Number Auto Generated Aadhaar Number User Name Aadhaar/User Age Aadhaar/User Gender Aadhaar Mobile OTP Verification Emergency Contact User State User Arrival Date User Transport Mode User Stay Type User Registration Time System Status Approved QR Code System ⸻ Validation Rules * Aadhaar number must be unique. * Aadhaar must pass checksum validation. * Mobile number must be OTP verified. * Arrival date cannot be in the past. * Emergency contact cannot match the primary mobile number (recommended validation). * All mandatory fields must be completed before registration. ⸻ Pass Number Format KM2027-000001 KM2027-000002 KM2027-000003 Generated sequentially using a database sequence or auto-increment mechanism. ⸻ QR Code Payload The QR code should encode a secure payload such as: { "pass_no": "KM2027-000001", "name": "Rajesh Kumar", "arrival": "2027-01-18", "status": "Approved" } For enhanced security, consider encoding only the pass number and retrieving visitor details from the server after scanning. ⸻ SMS / WhatsApp Confirmation Government of India Dear Rajesh Kumar, Your Kumbh Mela 2027 registration has been approved. Pass Number: KM2027-000001 Arrival Date: 18-Jan-2027 Please carry your Aadhaar Card and this pass during your visit. Download your pass: https://kumbh.gov.in/pass/KM2027-000001 ⸻ Recommended Technology Stack * Frontend: Next.js (responsive web application) * Backend: Node.js (NestJS or Express) or Java Spring Boot * Database: MySQL * Authentication: OTP via mobile number * Identity Verification: UIDAI Aadhaar eKYC (where legally authorized) * Voice AI: OpenAI Realtime API or Twilio ConversationRelay integrated with Deepgram/Whisper for speech-to-text and a natural TTS engine * SMS/WhatsApp: Government-approved SMS gateway and WhatsApp Business API * QR Code: Generated server-side and embedded in the digital pass * Hosting: NIC Cloud, AWS, Azure, or another government-approved cloud platform * Security: HTTPS, encrypted Aadhaar storage/tokenization, audit logs, role-based access control, and compliance with applicable Indian data protection and UIDAI regulations. This design supports both self-service web registration and AI-assisted phone registration, with both channels using the same backend APIs to ensure consistent validation, pass generation, and record management.