Customer Feedback Analysis (CFA)

An AI-powered customer feedback intelligence platform for Esme Beauty & Skincare. Ingests feedback from multiple sources (Amazon, Nykaa, social media, customer care), performs automated AI-driven sentiment analysis and complaint categorization, and surfaces actionable insights through an interactive dashboard and AI chat interface.

---

🚀 Key Features

• AI-Driven Data Ingestion — Automatic sentiment analysis and categorization using Google Gemini 2.0 Flash.
• Multi-source Support — Upload CSV/XLSX files with smart column auto-detection.
• Unified Analytics — Single source of truth for all customer feedback across platforms.
• AI Chat Agent (Esme) — Natural language queries over live feedback data with function calling for precise SQL analytics.
• Semantic Search — Find relevant feedback using AI-powered keyword and vector matching.
• Interactive Dashboard — KPI cards, monthly trends, and product-level deep dives.
• Dark Mode — Modern, responsive UI with GSAP animations.

---

🛠️ Tech Stack

Backend

• Runtime: Node.js
• Framework: Express.js
• Database: PostgreSQL (with pgvector support)
• AI Models:
  • gemini-2.0-flash (Sentiment & Categorization)
  • text-embedding-004 (Semantic Search)
• File Parsing: xlsx, csv-parse

Frontend

• Framework: React 18 (Vite)
• Styling: Tailwind CSS
• Charts: Recharts
• Animations: GSAP
• HTTP Client: Axios

---

📁 Project Structure

cfa/
├── backend/
│   ├── server.js           # Express app (Port 5000)
│   ├── db.js               # PostgreSQL connection
│   ├── schema.sql          # Database schema (feedback.complaints)
│   ├── database_setup.js   # Setup script
│   ├── routes/
│   │   ├── upload.js       # File upload & AI ingestion
│   │   ├── analytics.js    # Dashboard & Product APIs
│   │   └── chat.js         # AI Agent with Gemini
│   └── services/
│       └── processData.js  # AI Sentiment & Categorization pipeline
└── frontend/               # React + Vite app

---

🚦 Getting Started

1. Database Setup

Ensure PostgreSQL is running, then create the database and run the setup script:

psql -U postgres -c "CREATE DATABASE customer_feedback;"
cd backend
npm install
# Configure .env first (see below)
node database_setup.js

2. Environment Configuration

Create a .env file in the backend/ directory:

PORT=5000
PG_HOST=localhost
PG_PORT=5432
PG_USER=postgres
PG_PASSWORD=your_password
PG_DATABASE=customer_feedback
GEMINI_API_KEY=your_gemini_api_key

3. Run the Application

Backend:

cd backend
node server.js

Frontend:

cd frontend
npm install
npm run dev

Visit http://localhost:5173/cfa/

---

📊 AI Analysis Categories

When data is uploaded, Esme AI automatically classifies complaints into:

• Packaging Issue
• Health/Allergy Issue
• Counterfeit Concern
• Product Performance
• Product Quality
• Smell Issue
• Damaged Product
• Late Delivery
• Wrong Item
• Printing & Labeling
• Customer Service
• Other

---

🤝 Branding

Primary color: #0d968b (Teal) — Esme Beauty & Skincare. Developed for internal data intelligence and customer experience optimization.