OneCard SmartAssist

OneCard SmartAssist is a GenAI‑powered credit card assistant (React + FastAPI) that lets users manage their OneCard via chat and voice — check balance, view transactions, download statements, pay bills via QR/UPI, and track or block the card.

demo video link:-https://drive.google.com/drive/folders/16NH6icAv-hCsxNHZdBbyCoUpEUzCfV9l

Features

• Chat + voice interface with a virtual credit‑card agent
• Account overview: credit limit, used amount, available credit, and current bill
• Recent transactions list with merchant, date, and amount
• PDF statement download generated from mock data
• Bill payment via UPI QR and simple confirmation flow
• Spending analytics dashboard for category‑wise spends
• Card controls: track card delivery, block card with reason + confirmation
• Text‑to‑speech (TTS) for bot replies
• Optional speech‑to‑text (STT) via /speech-to-text + Whisper model

---

Architecture (High Level)

• React frontend

  • Screens: Register, Login, Chatbot, Dashboard
  • Components: Chatbot, Dashboard, StatementPDF
  • Handles quick intents on the client (balance, bill, transactions, block, etc.)
  • Sends API requests for LLM chat and speech‑to‑text

• FastAPI backend

  • REST endpoints for auth, chatbot streaming, and STT
  • Calls Llama‑based / OpenAI‑compatible LLMs for open‑ended queries
  • Calls Whisper‑style STT for voice input
  • Reads mock data and knowledge‑base text files for OneCard‑specific answers

• Data & knowledge

  • Demo user, card, bill, and transactions stored in local DB / mock data
  • Simple text files for FAQs: onboarding, delivery, EMI, payments, statements

---

Tech Stack

Frontend

• React (SPA)
• React Router
• react-qr-code for UPI QR
• @react-pdf/renderer for PDF statement
• Browser Web Speech API for TTS

Backend

• FastAPI
• Llama / OpenAI‑compatible LLM API (chat)
• Whisper‑style STT API
• CORS middleware
• SQLite / mock DB + text knowledge‑base

---

Getting Started

1. Clone the repo

git clone https://github.com/Deepa2410-ux/OneCard-SmartAssist.git cd OneCard-SmartAssist

2. Backend setup

cd backend python -m venv venv venv\Scripts\activate # Windows pip install -r requirements.txt

Create a .env file in backend/: OPENAI_API_KEY=your_llm_or_groq_key WHISPER_API_KEY=your_whisper_key # or same key if shared

Run FastAPI: uvicorn app:app --reload

Backend will be available at http://localhost:8000.

3. Frontend setup

cd ../frontend npm install npm start

Frontend will be available at http://localhost:3000.

---

Demo Flow

1. Register with demo details (name, phone, email, UPI PIN, last 4 digits).
2. Login using phone number and PIN.
3. Use quick‑action buttons:
   • Check Balance
   • Recent Transactions
   • Download Statement
   • Spending Analytics
   • Track Card
   • Block Card
   • Pay Bill
4. Click the mic icon to speak; the app:
   • Records your voice
   • Sends audio to /speech-to-text
   • Uses the transcribed text in the same chat flow

---

‘AI Co‑Pilot’ Usage

• Llama / OpenAI‑compatible LLMs power natural chat responses and fallback answers.
• Whisper‑style STT converts microphone audio to text for voice chat.
• Perplexity AI was used during development for researching issues, refining architecture, and drafting documentation/content, with all final design and code decisions reviewed manually.

---

Roadmap

• Real backend integration with live OneCard APIs
• Multi‑language support (English + Indian languages)
• Richer analytics and personalized insights
• Dispute / support ticket flows from within the chat

---