Author - LATHEESHAN THIRUCHANDURAN University - UNIVERSITY OF MORATUWA Year- 3rd Year DEMO VIDEO PITCH (DRIVE LINK) - https://drive.google.com/drive/folders/18JDLC2nvdR85czxg4o0ivf_PQpYSlhbL?usp=sharing

      Drive folder includes working vido demonstration + interface Images 

       https://drive.google.com/drive/folders/18JDLC2nvdR85czxg4o0ivf_PQpYSlhbL?usp=sharing

This project is a multi-agent feedback analysis and response system built with FastAPI (backend) and Next.js (frontend). It uses LangChain / LlamaIndex with LLMs to automate customer sentiment analysis and feedback response generation.

The system is designed to:

1. Agent 1 – Customer Feedback Response Agent

   • Accepts customer feedback as input.
   • Uses an LLM to classify sentiment (Positive / Negative / Neutral).
   • Generates a short, polite, and context-aware automated reply.

2. Agent 2 – Sentiment Visualization Agent

   • Accepts a date range (e.g., “last 7 days”, or “June 1 to June 15”).
   • Dynamically generates bar/line plots showing the number of positive, negative, and neutral reviews per day.

• Backend: FastAPI, Uvicorn, Pydantic, Pymongo, Requests
• Frontend: Next.js (React framework)
• Frameworks: LangChain / LlamaIndex
• LLMs: HuggingFace Transformers, Groq LLM (via langchain_groq)
• Python Libraries: pandas, matplotlib, plotly
• Dataset: defind script backend

#backend/ │── controllers/ # API route handlers │ ├── feedback_controller.py │ ├── feedback_trend_controller.py │ │── models/ # Database models │ ├── feedback_model.py │ │── schemas/ # Pydantic schemas │ ├── feedback_schema.py │ │── services/ # Business logic & AI services │ ├── ai_sentiment.py │ ├── feedback_service.py │ │── venv/ # (Optional) Virtual environment │ │── auth.py # Authentication logic │── config.py # Configuration settings │── database.py # MongoDB connection │── main.py # FastAPI entry point │── mock_feedback_data.py # Mock data for testing │── requirements.txt # Project dependencies │── .env # Environment variables

#frontend/ │── public/ # Static assets (images, icons, etc.) │── src/app/ # App Router pages │ ├── feedback/visualization/ # Feedback visualization pages │ │ ├── page.tsx │ ├── globals.css # Global styles │ ├── layout.tsx # Root layout │ ├── page.tsx # Landing page │ │── .gitignore │── eslint.config.mjs # ESLint configuration │── next-env.d.ts # Next.js type definitions │── next.config.ts # Next.js configuration │── package.json # Project dependencies │── package-lock.json # Lockfile │── postcss.config.mjs # PostCSS configuration │── tsconfig.json # TypeScript configuration │── README.md # Documentation

🛠️ Full Project Setup & Running Guide

PREREQUSTS

✅ Prerequisites for Project

🔹 General Tools (must have installed)

Git → for cloning and version control

Python 3.9+ → required for FastAPI + LangChain

Node.js 18+ and npm (or yarn) → required for Next.js frontend

MongoDB (local or cloud via MongoDB Atlas) → for storing feedback data

Virtual Environment (venv) → to manage Python dependencies

1️⃣ Clone the Repository git clone cd

2️⃣ Backend Setup (FastAPI) Navigate to backend folder cd backend

Create & activate virtual environment (recommended)

python -m venv venv

venv\Scripts\activate

source venv/bin/activate

#Install dependencies

pip install -r requirements.txt

#Setup .env file

NOTE- ALREADY BACKEND HAS THE .env FILE , IF IT'S MISSING CREATE A NEW .env FILE AND PASTE THIS BELOW DATA

MONGODB_URI=mongodb://localhost:27017 DB_NAME=feedback GROQ_API_KEY=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

#Run mock data script This will generate or load your mock data into memory / file / DB (depending on your implementation).

python mock_feedback_data.py

#SAMPLE DATA IN MONGODP LOOKS LIKES THIS

{ "_id": "d44f3bd6-1f7f-47b5-ae9d-ff06a05370f5", "text": "This is a neutral feedback example.", "sentiment": { "label": "neutral", "score": 0.77, "response": "Thanks for your neutral feedback!" }, "createdAt": { "$date": "2025-08-15T13:03:17.838Z" } }

#Start backend server

Once mock data is ready, run:

uvicorn main:app --reload

3️⃣ Frontend Setup (Next.js) Open a new terminal & navigate to frontend

cd frontend

#Install dependencies

npm install

#Run the frontend

npm run dev

4️⃣ Running Together

#Start backend (mock API):

cd backend uvicorn main:app --reload

#Start frontend (in another terminal):

cd frontend npm run dev

Open in browser: 👉 http://localhost:3000

Frontend will now fetch mock feedback data from the backend API.