🤖 Personality Chatbot Server

A Node.js backend chatbot powered by Groq LLM with a unique personality system, memory, and randomness.

---

🚀 Features

• 🎭 Strong personality (emoji + pauses + smart/dumb behavior)
• 🧠 Short-term + long-term memory (MongoDB)
• 👽 Random alien / elvish / conflict events
• ⚡ Fast responses using Groq (free LLM API)
• 🔒 Secure environment variable handling

---

🛠️ Tech Stack

• Node.js
• Express.js
• MongoDB (Mongoose)
• Groq API (Free LLM)

---

📦 Installation & Setup

1. Clone the repository

git clone https://github.com/aditig80/chatbot-server
cd chatbot-server

2. Install dependencies

npm install

3. Create .env file

Copy from .env.example:

PORT=5000
GROQ_API_KEY=your_api_key_here
MONGO_URI=your_mongodb_connection_string

---

▶️ Running the Server

node server.js

Expected output:

Server running on port 5000
MongoDB connected

---

🧪 Testing with Postman

Endpoint

POST http://localhost:5000/chat

Request Body (JSON)

{
  "userId": "user1",
  "message": "How do I prepare for interviews?"
}

Response Example

{
  "reply": "💡 You should focus on DSA fundamentals ||| practice mock interviews ||| build projects"
}

---

📸 Screenshots

1. Postman Request - Response

2. MongoDB Data

---

🧠 System Prompt Design

The chatbot personality is controlled using a structured system prompt with strict rules:

Key Rules

1. Emoji Rule

   • Every response starts with exactly one emoji.

2. Pause Rule

   • Uses ||| as natural pauses.

3. Dual Intelligence Behavior

   • Career topics → smart, structured, helpful
   • Other topics → dumb, confused, humorous

4. Random Behaviors

   • Alien language glitches
   • Elvish phrases
   • Mentions of elf vs alien conflict

Why this structure?

• Ensures consistent personality
• Creates contrast (smart vs dumb)
• Makes chatbot feel unique and alive
• Backend logic enforces rules to avoid LLM inconsistency

---

🧠 Memory Implementation

Short-Term Memory

• Stores last 10 messages per user
• Maintains conversation context

Long-Term Memory (MongoDB)

• Each user has a document:

{
  "userId": "user1",
  "messages": [
    { "role": "user", "content": "Hello" },
    { "role": "assistant", "content": "..." }
  ]
}

How it works

1. Fetch conversation from DB
2. Append new messages
3. Trim to last 10 messages
4. Save back to MongoDB

Benefits

• Persistent memory across sessions
• Context-aware replies
• Scalable for multiple users

---

📁 Project Structure

chatbot-server/
│
├── server.js
├── package.json
├── .env.example
├── README.md
├── screenshots/

---

🚀 Future Improvements

• Rate limiting
• Deployment (Render / Railway)
• Logging system
• UI frontend (optional)

---

👩‍💻 Author

Aditi Gupta