Theme 1

15-Day GenAI Learning Challenge (Beginner Edition)

From Zero → Your First AI App in 2 Weeks

---

Design Principles (Important)

• Beginner-safe: No deep math, no ML rabbit holes

• Hands-on first: Every day produces something tangible

• Modern GenAI stack: Python + LLMs + RAG

• Portfolio-ready: End result is demo-worthy

---

Day 1 – Environment & Mindset

Goal: Become “AI-ready” in one sitting

• Install Python 3.10+ (Conda or system Python)

• Install VS Code

• Create virtual environment

• Run first Python script

• Understand:

  • What GenAI is

  • Difference between ML vs GenAI vs LLMs

Output: ✔ Python installed ✔ hello_genai.py

---

Day 2 – Python Essentials for GenAI

Goal: Learn only the Python you actually need

• Variables, lists, dicts

• Functions

• Reading & writing files

• JSON basics

GenAI relevance: LLMs talk in JSON, configs, prompts

Output: ✔ Python script that reads a text file and processes it

---

Day 3 – Git & Developer Workflow

Goal: Learn how real GenAI engineers work

• Git basics: init, commit, push

• GitHub repo creation

• README writing (simple)

Output: ✔ First public/private GitHub repo ✔ README.md

---

Day 4 – What Is an LLM (No Math)

Goal: Understand LLMs conceptually without confusion

• What is a Large Language Model

• Tokens, context window, temperature

• Prompt → Response mental model

• Cloud LLM vs Local LLM (Ollama)

Output: ✔ Clear mental model (most beginners skip this and fail)

---

Day 5 – Your First LLM Call

Goal: Talk to an AI using code

• Choose OpenAI API or Local Ollama

• Install SDK

• Send prompts from Python

• Adjust temperature and max tokens

Output: ✔ Python script that asks questions to an LLM

---

Day 6 – Prompt Engineering Basics

Goal: Learn why prompts fail and how to fix them

• Zero-shot vs Few-shot

• System vs User prompts

• Structured prompts

• Prompt templates

Output: ✔ Prompt that consistently produces good answers

---

Day 7 – Mini Project: AI Text Assistant

Goal: Build confidence

• CLI-based AI assistant

• Input → LLM → Output

• Save conversations to file

Output: ✔ Working AI assistant (terminal-based)

---

Day 8 – Embeddings (The Magic Behind Search)

Goal: Understand how AI “remembers”

• What embeddings are

• Why similarity search works

• Convert text → vectors

• Cosine similarity (conceptual)

Output: ✔ Script that converts text into embeddings

---

Day 9 – Vector Database (Beginner Version)

Goal: Store and retrieve knowledge

• FAISS or Chroma (local)

• Insert documents

• Query similar content

Output: ✔ Vector DB with your own documents

---

Day 10 – RAG: Retrieval-Augmented Generation

Goal: Build your first real GenAI system

• Why LLMs hallucinate

• How RAG fixes hallucinations

• Flow:

  • User query

  • Vector search

  • Context injection

  • LLM answer

Output: ✔ AI that answers from your documents

---

Day 11 – RAG Mini Project

Goal: Apply RAG end-to-end

• Upload PDFs or text files

• Ask questions

• Get grounded answers

Output: ✔ Personal document Q&A bot

---

Day 12 – FastAPI: Make It Real

Goal: Turn scripts into an app

• What is FastAPI

• Simple /ask endpoint

• Send prompt via API

• Return AI response

Output: ✔ Running GenAI API locally

---

Day 13 – AI Agent Basics

Goal: Introduce autonomy (without hype)

• What is an AI agent

• Tool-calling concept

• Agent vs chatbot

• Simple “decide → act → respond” loop

Output: ✔ Simple rule-based AI agent

---

Day 14 – Final Project

Goal: Combine everything

Build one of the following:

• AI Study Assistant

• Resume Reviewer

• Personal Knowledge Bot

• PDF Chatbot

Must include:

• LLM

• Embeddings

• RAG

• API

Output: ✔ End-to-end GenAI application

---

Day 15 – Packaging & Showcase

Goal: Make it portfolio-ready

• Clean README

• Architecture diagram (simple)

• Demo screenshots

• What you learned

Output: ✔ Shareable GenAI project ✔ Confidence to continue deeper

---

What This 15-Day Challenge Achieves

By Day 15, a beginner will:

• Understand how GenAI actually works

• Write real GenAI code

• Build a RAG-based AI app

• Avoid 90% of beginner mistakes

• Be ready for:

  • LangChain

  • LangGraph

  • Multi-agent systems

  • Enterprise GenAI

---