Weekend Projects
Build Something Real — By Monday
5 hand-picked GitHub repos. Python, data science, ML, MLOps, RAG, AI agents. Pick one Friday night, ship by Sunday. No more bookmarks that rot.
Finished a weekend project? Build a production-grade Agentic RAG system next.
7 weeks · 12 hours · Self-paced · Live Now · Limited time
JMJCAWJ64E The 5 Projects
Every repo here is one I'd actually recommend to a friend.
Different paths, different vibes. Pick the one that scares you the right amount.
Project-Based Learning
practical-tutorials/project-based-learningThe definitive list of build-by-doing tutorials. Build your own BitTorrent client, shell, git, neural net, interpreter, 3D renderer — in Python, Go, Rust, JS, and more. No theory walls. You open the repo, pick a project, follow the article, ship.
What you'll build
- Your own text editor, web server, or database
- A neural network from scratch — forward and back prop
- A shell, interpreter, or compiler in your language of choice
# Pick this weekend if…
- You keep watching tutorials but never finish anything
- You want to learn Python by building, not reading
- You want 100+ ideas in one place you can trust
$ clone it
git clone https://github.com/practical-tutorials/project-based-learning
Beginner Data Science Projects
tkarim45/Beginner-Data-Science-ProjectsA tight collection of beginner-friendly data science projects with datasets included. Titanic survival, house prices, customer churn, sentiment analysis. Each one walks through cleaning, EDA, modeling, and evaluation — the full loop, not just the fun part.
What you'll build
- End-to-end EDA → modeling → evaluation pipelines
- Clean Jupyter notebooks worth putting on GitHub
- The habit of documenting your reasoning, not just your code
# Pick this weekend if…
- You finished a course and need your first real project
- Your GitHub is empty and you're applying for DS roles
- You want to practice the full workflow, end to end
$ clone it
git clone https://github.com/tkarim45/Beginner-Data-Science-Projects
MLOps Coding Course
MLOps-Courses/mlops-coding-courseStop training models that die in notebooks. This repo teaches you to structure an ML project like real teams do — poetry, pre-commit, pytest, MLflow, CI/CD, Docker, the whole spine. You'll feel the difference between a Kaggle notebook and a shipped system.
What you'll build
- A fully-structured ML project with testing and packaging
- CI/CD pipelines for models — not just code
- Experiment tracking, model registry, and deployment flow
# Pick this weekend if…
- Your notebooks work but you've never shipped one
- You want to earn 'ML Engineer' and not just 'Data Scientist'
- You're tired of re-running cells to reproduce your results
$ clone it
git clone https://github.com/MLOps-Courses/mlops-coding-course
Beginner Local RAG System
jamwithai/beginner-local-rag-systemA complete RAG system you can run on your laptop — no OpenAI key, no cloud bill. Chunking, embeddings, a vector store, retrieval, and an LLM answer loop, all with local models. The clearest way to actually understand what's happening inside every RAG tutorial you've watched.
What you'll build
- A working local RAG pipeline with Ollama + a vector DB
- Chunking, embedding, and retrieval you can reason about
- A mental model for every production RAG you'll build after
# What you'll actually build
Ingest (one-time)
docs/
PDFs · MDs
chunk
split + overlap
embed
Ollama · local
vector store
ChromaDB · runs local
Query (every request)
?
question
retrieve
top-k chunks
LLM
Ollama · answer
500 AI Agents Projects
ashishpatel26/500-AI-Agents-ProjectsFive hundred agent project ideas across industries — finance, health, legal, research, ops, gaming. Each entry has a use case, the agent's role, and which tools it would need. When you're staring at a blank screen wondering 'but what should I build?' — this repo is the cure.
What you'll build
- Pick an idea, scope it down, ship an MVP in a weekend
- A small portfolio of 2–3 agents in domains you care about
- Pattern recognition for what makes a good agent problem
# Pick this weekend if…
- You know how to build agents but have no ideas
- You want industry-specific inspiration, not more chatbots
- You're building a portfolio and need breadth
$ clone it
git clone https://github.com/ashishpatel26/500-AI-Agents-Projects
The Hard Part
How to actually finish
Picking the project is the easy part. These four habits are what separate a shipped repo from a stale branch.
Fork first, commit as you go
Your future self wants the history. Small commits with honest messages beat one giant 'final' push.
Write a README for yourself
Not for a hiring manager. Explain what you built and why, in your own voice. That's the portfolio piece.
Ship it rough, polish in v2
A shipped rough version beats a polished one that never makes it out of your local. v2 is allowed.
Post it before you're ready
Share the repo, share the lesson, share the thing you broke and fixed. That's where the loop closes.
Go Deeper
When the weekend isn't enough — build a production RAG
Finished the beginner RAG repo? This is the next step. Same architecture real teams ship: observability, hybrid retrieval, agentic workflows, streaming APIs.
Agentic RAG System
Build a Production-Ready RAG System from Scratch
12 hours of hands-on content. Build an arxiv paper curator with hybrid retrieval, LLM generation via Ollama, agentic workflows, full observability, and streaming APIs. 23+ tools, real architecture, no hand-waving.
- Vector databases & embedding models
- Hybrid search with BM25 + RRF
- Monitoring, caching & evaluation
- Agentic RAG with LangGraph
- Streaming APIs & FastAPI backend
- Ollama for local LLM inference
Self-paced · 12 hours · 7 weeks
JMJCAWJ64E Not ready for the course? Stay in the loop.
Weekly drops of resources like this, behind-the-scenes on what we're shipping, and early access to new courses.