🎯 Customer Relationship Management System

A full-stack Customer Relationship Management System where admins can manage customers, segment them using advanced rules or AI text input, and create personalized campaigns with communication logs and AI-generated summaries.

🌐 Deployment

🔗 Live CRM App

🚀 Live Demo

📹 Video Walkthrough

🧭 Flowchart

---

🔁 Flow Summary

1. 🔐 Login Route (/login)

• Admin logs in via Google, GitHub, or credentials (using Passport.js).

2. 👥 Add Customers (/customer)

• Admin can:
  • Add customers via a form
  • Add orders per customer via a modal popup
  • View customer profile and order in a quick popup

3. 🧩 Create Segments (/segment)

• Two options:

  • A rule based form to build customer segments:

    • Fields: daysInactive,totalSpend,numberOfVisits,accountType,lastPurchaseDate,isSubscribed,mostCategoryOfProductsPurchased
    • Flexible AND/OR logic
    • Matching customers can be previewed with total count

  • ✨ AI text area to describe desired segment in natural language → generates JSON rules

    • Matching customers can be previewed with total count

4. 💬 Campaign Messaging

• Two options:
  • ✨ AI-generated message suggestions
  • ✍️ Personalized Campaign messages suggestions by inputting topic and description
• On submission:
  • Campaign is saved
  • Communication logs created per customer
  • Message sent → status updates (pending → sent/failed)

5. 📈 View Campaigns (/campaign)

• Displays:
  • Campaign details & message
  • Delivery stats (from communication log)
  • AI-generated summary, e.g.:

    “Your campaign reached 1,200 users. 1,050 were successfully delivered. 150 failed. Customers who spent more than ₹10K had the highest success rate.”

---

🧠 AI Tools Used

Tool	Purpose
Google Generative AI	Campaign summaries, segment rule generation, personalized message suggestions

---

📸Demo Images

💡 Smart Segmentation UX

Avoided cluttering with too many AND/OR dropdowns.
Used a rule-builder UI: React Query Builder

✅ Cleaner UX
✅ Nested conditions
✅ Easy to understand

---

⚠️ Known Limitations

• Message delivery is simulated (no real messaging API)
• No retry mechanism for failed messages
• AI generation depends on quality of text input

---

🛠️ Local Setup

Prerequisites

• Node.js (v18+)
• MongoDB (local or Atlas)
• npm

Clone the Repository

# Backend
cd backend
npm install
npm start

# Frontend
cd frontend
npm install
npm run dev

---

🔹 Frontend Stack

Package	Purpose
Next.js	As a Frontend Framework
React,ReactDom	Next is build on React
tailwind css	Styling
Shadcn	Enhanced UI
axios	API calls
react-hot-toast	Toast notifications
react-querybuilder	Advanced rule builder UI

---

🔹 Backend Stack

Package	Purpose
Express.js	Web server
Mongo DB	MongoDB
passport, passport-local, passport-google-oauth20, passport-github2	Authentication strategies
express-session	Session management
jsonwebtoken, bcrypt	JWT + password hashing
Google Gemini API	AI integration
dotenv, cors	Config and CORS

Backend Deployed Url

https://connect-crm-backend.onrender.com/

Frontend Deployed Url

https://xeno-crm-wo6w.onrender.com/