Skip to content

README.md

Latest commit

 

History

History
77 lines (50 loc) · 1.67 KB

README.md

File metadata and controls

77 lines (50 loc) · 1.67 KB
Raw

Vittam

Vittam is a multi-agent personal loan sales system designed to streamline the loan application process. It leverages AI agents to handle various stages of the loan journey, including sales, verification, underwriting, and sanctioning.

Features

  • Master Agent: Orchestrates the entire loan journey, handles customer needs, and routes tasks to the appropriate agents.
  • Sales Agent: Understands customer needs, handles objections, and generates personalized loan offers.
  • Verification Agent: Performs KYC, PAN, and phone verification.
  • Underwriting Agent: Assesses credit scores, eligibility, and calculates EMIs.
  • Sanction Agent: Generates sanction letters and provides disbursement details.

Prerequisites

  • Python 3.13+
  • Node.js (for client-side components)
  • MongoDB (for database operations)

Setup Instructions

Backend

  1. Install dependencies:

    pip install -r requirements.txt

  2. Set environment variables in the .env file

  3. Run the server:

    uv run runner.py

Frontend

  1. Navigate to the client directory:

    cd client

  2. Install dependencies:

    pnpm install

  3. Start the development server:

    pnpm dev

Chatbot

  1. Navigate to the chatbot directory

  2. Install dependencies:

    pnpm install

  3. Run the chatbot:

    nodemon

Project Structure

  • server/: Backend services and AI agent logic.
  • client/: Frontend React components for user interaction.
  • chatbot/: The interactive chatbot code and logic