Cognexa AI : LLM-Powered AI Mentor System

Cognexa-AI is a modular, LLM-powered AI Mentor System designed to provide subject-specific mentoring with a professional, industry-grade approach. The system simulates a senior technical mentor with 15+ years of MNC experience.

---

Project Vision

The goal of Cognexa-AI is to build a domain-restricted, mentor-style AI chatbot that delivers:

• Clear and structured explanations
• Industry-relevant examples
• Encouraging and professional mentoring tone
• Strict subject-based responses (no hallucination)

---

📁 Project Structure

AI_Chatbot_mentor/
│
├── notebook/
│   ├── app.py        # Final Streamlit application (UI-focused)
│   ├── flow.py       # Core architecture & logic of the chatbot
│   ├── .env          # Environment variables (API keys)
│   └── req.txt       # Project dependencies
│
├── Include/
├── Lib/
├── Scripts/
├── share/
└── pyvenv.cfg

---

🧠 flow.py — Core Logic & Architecture

flow.py contains the main structural and logical backbone of Cognexa-AI.

Think of flow.py as the brain of the system.

Responsibilities:

• Application flow control (Home → Module → Chat)
• Module-wise subject selection (Python, SQL, ML, etc.)
• Prompt engineering using ChatPromptTemplate
• LLM streaming integration (Gemini)
• Conversation memory using st.session_state
• Strict domain restriction logic
• Chat history persistence
• Conversation download feature

---

🎨 app.py — Final Application & UI Layer

app.py represents the final, user-facing application.

app.py focuses on user experience, not core logic.

Responsibilities:

• Clean and professional Streamlit UI
• Layout, headings, icons, and spacing
• User-friendly navigation
• Deployment-ready application

Summary:

• flow.py → How the system works
• app.py → How the system feels

---

🔐 .env — Environment Variables

The .env file stores sensitive credentials securely.

Gemini=YOUR_GOOGLE_GEMINI_API_KEY

• Keeps API keys out of source code
• Improves security
• Must be added to .gitignore

---

📦 req.txt — Dependencies

The req.txt file lists all required Python packages.

streamlit
langchain
langchain-google-genai
python-dotenv

Install dependencies using:

pip install -r req.txt

---

🧩 Design Philosophy

• Logic and UI are cleanly separated
• Prompt engineering is centralized
• Each module has a dedicated mentor
• Production-style architecture inspired by real MNC systems

---

✅ Why Cognexa-AI Stands Out

• Portfolio-ready project
• Internship and placement friendly
• Scalable and extensible design
• Professional AI mentoring experience

Chatbot.demo.mp4

---

✨ Built with passion, structure, and industry mindset.