Theme2

100 Days of GenAI Challenge — Daily Tasks


Days 1–10: Python + Conda Setup

  • Day 1: Install Miniconda/Anaconda, Python 3.11

  • Day 2: Create your Conda environment and activate it

  • Day 3: Install Jupyter Notebook, create your first Notebook

  • Day 4: Python Basics: Variables, Data Types, Input/Output

  • Day 5: Python Basics: Functions, Conditionals, Loops

  • Day 6: Python Basics: Lists, Dictionaries, Tuples, Sets

  • Day 7: Python Basics: Classes and Object-Oriented Programming

  • Day 8: Setup Git + GitHub, create your first repository

  • Day 9: Write a README.md with Markdown for your first repo

  • Day 10: Share your setup journey on LinkedIn


Days 11–20: Core Machine Learning Basics

  • Day 11: Install pandas, numpy, matplotlib

  • Day 12: Load and clean a dataset (CSV) using pandas

  • Day 13: Build a simple Linear Regression model (e.g., house prices)

  • Day 14: Visualize data (scatter plots, histograms)

  • Day 15: Train/Test Split - Understand overfitting/underfitting

  • Day 16: Classification: Train a Decision Tree classifier

  • Day 17: Evaluate models: Accuracy, Confusion Matrix

  • Day 18: Build a simple Machine Learning Pipeline

  • Day 19: Save model using joblib

  • Day 20: Share your first ML project on LinkedIn


Days 21–30: Language and LLM Fundamentals

  • Day 21: Introduction to NLP (Text preprocessing)

  • Day 22: Install and explore Huggingface Datasets

  • Day 23: Tokenization basics (word-level and subword)

  • Day 24: Install and use OpenAI API for simple prompts

  • Day 25: Understand Prompt Parameters: temperature, top_p, max_tokens

  • Day 26: Few-shot Prompting: design your own prompt examples

  • Day 27: Chain of Thought Prompting — Reasoning examples

  • Day 28: Install and load GPT2 using Huggingface Transformers

  • Day 29: Compare GPT2 generated text vs OpenAI API text

  • Day 30: Write a blog about "Prompting My First Language Models"


Days 31–40: Embeddings and RAG (Retrieval-Augmented Generation)

  • Day 31: What are Embeddings? Install OpenAI Embeddings

  • Day 32: Generate sentence embeddings and visualize with PCA

  • Day 33: Install FAISS, index your embeddings locally

  • Day 34: Perform semantic search on your FAISS index

  • Day 35: Load a PDF document, split into chunks

  • Day 36: Generate embeddings for document chunks

  • Day 37: Create a basic RAG system (Search + LLM Answer)

  • Day 38: Try Huggingface Embeddings instead of OpenAI

  • Day 39: Experiment with Chroma or Qdrant as vector store

  • Day 40: Publish a post: "I Built My First RAG Bot!"


Days 41–50: LangChain Basics

  • Day 41: Install LangChain

  • Day 42: Build your first simple LangChain LLM Chain

  • Day 43: Create a Sequential Chain (multi-step task)

  • Day 44: Add Memory to your chain (ConversationBufferMemory)

  • Day 45: Build a Question-Answering Chain using documents

  • Day 46: Create your own Custom Prompt Templates

  • Day 47: Create your first Tool and use in LangChain Agent

  • Day 48: Chain Tools into a mini-Agent

  • Day 49: Integrate LangChain Agent + FAISS for smart search

  • Day 50: Share your Agent project on LinkedIn


Days 51–60: Tool-Using Agents

  • Day 51: Introduction to Tools and Agents (high-level)

  • Day 52: Add Search API as a Tool

  • Day 53: Add Calculator API as a Tool

  • Day 54: Create a multi-tool agent that decides what to use

  • Day 55: Understand "tool selection" based on queries

  • Day 56: Write custom Output Parsers for LangChain

  • Day 57: Create a Financial Analyst Agent (search + calculate)

  • Day 58: Create a simple Email Generator Agent

  • Day 59: Implement Logging and Error Handling in your Agents

  • Day 60: Publish: "How I Built a Multi-Tool AI Agent"


Days 61–70: Autonomous Agents (AutoGPT/BabyAGI Inspired)

  • Day 61: Study how BabyAGI works (repo exploration)

  • Day 62: Build a simple Task Creation -> Execution loop

  • Day 63: Add memory store to keep track of completed tasks

  • Day 64: Add prioritization logic

  • Day 65: Self-evaluate task quality

  • Day 66: Replan based on new inputs (dynamic replanning)

  • Day 67: Create a "Daily Content Generator" autonomous agent

  • Day 68: Limit API usage to optimize costs

  • Day 69: Debugging Autonomous Flows

  • Day 70: Share your experience: "My First Autonomous Agent"


Days 71–80: FastAPI + Docker for GenAI Apps

  • Day 71: Install and learn FastAPI Basics

  • Day 72: Build a simple FastAPI endpoint for OpenAI API

  • Day 73: Serve your RAG app as a FastAPI service

  • Day 74: Docker Basics: Containers, Images, DockerHub

  • Day 75: Write a Dockerfile for your FastAPI app

  • Day 76: Build and run your container locally

  • Day 77: Push your app to DockerHub

  • Day 78: Deploy your container to a free cloud (Render, Fly.io)

  • Day 79: Monitor API logs and handle exceptions gracefully

  • Day 80: Publish: "Deploying My First GenAI API to Cloud"


Days 81–90: Prefect Orchestration for GenAI Workflows

  • Day 81: Prefect Basics: Flows and Tasks

  • Day 82: Create a Prefect flow for document processing

  • Day 83: Schedule your flow with Prefect Work Pools

  • Day 84: Build a Flow that runs every day: "Daily Tweet Writer"

  • Day 85: Add concurrency and retries to your Prefect flows

  • Day 86: Use environment variables securely (Prefect Blocks)

  • Day 87: Automate RAG Reindexing weekly

  • Day 88: Handle Prefect Failures with Notifications (Slack, Email)

  • Day 89: Prefect Deployment using Prefect YAML

  • Day 90: LinkedIn Post: "Automating GenAI Workflows using Prefect"


Days 91–100: Agentic Workflows, MCP, and A2A

  • Day 91: Introduction to MCP (Model Context Protocol)

  • Day 92: Understand StdioTransport, Server classes

  • Day 93: Install and run MCP Python SDK

  • Day 94: Build a simple MCP Server with one Tool

  • Day 95: Extend to Multi-Tool MCP Server

  • Day 96: Build an Agent that reads documentation and writes summary

  • Day 97: Build another Agent that reviews summaries and corrects them

  • Day 98: Connect Agents in A2A (Agent to Agent) style

  • Day 99: Showcase: Create a multi-agent autonomous publishing system

  • Day 100: Grand Finale Post:

"From Python Basics to Autonomous Agents: My 100 Days of GenAI Journey"


Extra:

  • Every 7–10 days: Do a LinkedIn Post summarizing your progress.

  • GitHub Public Portfolio: push code every week.

  • Use consistent hashtags: #100DaysOfGenAI | #LearnInPublic | #AIStudentChallenge.

Last updated