🚀 Gen AI Beginner Projects

Welcome to Gen AI Beginner Projects👋 - a hands-on learning repository designed to help you understand the fundamentals of Generative AI by building real-world projects using Python and modern LLMs.

This repository focuses on practical implementation of concepts like:

• Prompt Engineering
• LLM APIs (like OpenAI API)
• Embeddings
• Vector Databases
• RAG (Retrieval-Augmented Generation)
• AI Agents

Instead of just learning theory, this repo emphasizes learning by building.

If you’ve ever wondered:

• How does ChatGPT actually work?
• How can I build my own AI-powered app?
• How do companies use Generative AI in real products?

This repository is the practical answer to those questions.

🎯 Purpose of This Repository

This repository contains small, focused Gen AI projects, each living in its own folder, designed to:

• 📌 To build strong foundational knowledge in Generative AI
• 🛠️ To create practical, portfolio-ready projects
• 📚 To document experiments and learnings
• 🚀 To transition from beginner to intermediate level in Gen AI
• 💼 To prepare for AI/ML/LLM-based roles
• 🧠 To build intuition before production-grade systems

Each project is intentionally designed to focus on one or two core concepts to avoid overwhelming complexity.

⚠️ This repo is not production-ready.

🧠 What You’ll Learn from This Repo

By going through these projects, you will learn:

• Prompt Engineering
  • Zero Shot, Few Shot, Chain-of-Thought Prompting
  • System vs User Prompts
  • Temperature, Top-p tuning
  • Output Structuring (JSON, Markdown)
• Working with LLM APIs
  • Making API calls
  • Handling tokens & rate limits
  • Streaming response
  • Error Handling
• Embeddings & Semantic Search
  • What embeddings are
  • Converting text -> vectors
  • Cosine similarity
  • Semantic retrieval
• Vector Databases
  • Indexing embeddings
  • Storing & quering vectors
  • Similarity search
• RAG (Retrieval-Agumented Generation)
  • Chunking strategies
  • Context injections
  • Reducing hallucinations
  • Knowledge-grounded response
• AI Agents
  • Tool usage
  • Task planning
  • Multi-step reasoning
  • Basic agent workflows

🚀 Who Should Use This Repo?

This repository is perfect for:

• 🧑‍🎓 Students learning Generative AI
• 👨‍💻 Developers transitioning into AI
• 🔄 Professionals switching to AI/ML roles
• 🚀 Anyone who wants to build an AI portfolio
• 🧠 Curious minds who want to understand how modern AI works

If you are at beginner level and want structured, practical learning — this repo is for you.

Generative AI Learning Roadmap

graph TD

A[Start: Generative AI Basics] --> B[Prompt Engineering]
B --> C[LLM APIs]
C --> D[Build Simple Apps<br/>Summarizer / Email Generator]

D --> E[Embeddings]
E --> F[Semantic Search]
F --> G[Vector Databases]

G --> H[RAG - Retrieval Augmented Generation]
H --> I[Build Chat with PDF / Knowledge Bot]

I --> J[AI Agents Basics]
J --> K[Tool Calling]
K --> L[Multi-step Reasoning Apps]

L --> M[Intermediate Gen AI Projects]

Loading

📂 Repository Structure

gen-ai-beginner-projects/
│
├── ai-text-summarizer-gemini-python/
│   ├── gemini-text-summarizer.py
│   ├── READEME.md
│
├── ai-email-writer-gemini-python/
│   ├── gemini-email-writer.py
│   ├── READEME.md
│   
├── ...
│
├── requirements.txt
├── .env.example
├── .gitignore
└── README.md

🛠 Tech Stack

Common stack used across experiments:

• Language: Python 3.10+
• LLMs: Google Gemini
• Frameworks: langchain
• Env Management: python-dotenv
• IDE: VS Code (recommended)

⚙️ Environment Setup (One-Time)

Step 1: Clone the Repository

git clone https://github.com/pankaj-kaushik/gen-ai-beginner-projects.git
cd gen-ai-beginner-projects

Step 2: Create a Virtual Environment

python -m venv .venv

Activate it:

Windows (PowerShell):

.venv\Scripts\Activate.ps1

Mac/Linux:

source .venv/bin/activate

Step 3: Setup Environment Variables

Create a .env file at the root

GEMINI_API_KEY="your_api_key_here"
OPENAI_API_KEY="your_api_key_here"

👉 Refer to .env.example
👉 Never commit .env to GitHub

Step 4: Install python libraries

pip install -r requirements.text

▶️ How to Run A Project

Each project follows the same basic flow.

Step 1: Move into the project folder

cd ai-text-summarizer-gemini-python

Step 2: Run the python application

python gemini-text-summarizer.py

🧪 How to Add a New Project

Follow this standard structure:

project-name/
├── app.py
├── requirements.txt (if any)
├── README.md
└── assets/ (optional)

⚠️ Use similar file/folder naming convention

Best practices:

• Focus on one GenAI concept
• Keep code minimal & readable
• Document:
  • what the experiment does
  • what you will learn
  • how to run it

🧭 Planned / Example Experiments

Some experiments you may find or add

• Study Buddy / Concept Explainer
• Email / Message Writer
• Chat with PDF
• Resume Analyzer
• Prompt Comparison Lab
• Embeddings Search Demo
• RAG (Retrieval Augmented Generation)
• Simple AI Agent
• Interview Question Generator
• Career Coach For S/W Engineers
• Bug Root Cause Analyzer
• Teacher for kids (Math/Logic)

🎓 Learning Resources

• Google Gemini Documentation
• Prompt Engineering Guide
• Understanding Temperature & Top-P
• Python Best Practices

⚠️ Disclaimer

• LLM APIs may incur cost
• Outputs may vary due to model randomness
• This repo is for learning & experimentation only

🙌 Acknowledgements

• Google Gemini LLM
• Open-source Python community

Contributing

💡 If you found this helpful...

• ⭐ Star the repo
• 🍴 Fork it
• 🚀 Build on top of it & submit pull request
• 📢 Share your AI story platform

Happy Hacking 🧪🚀