Learn data science & AI by building full-stack projects
Full-Stack Data Science is a complete program that teaches you everything required for modern data science - Python, SQL, Git, ML, AI, and MLOps.
Data science is changing
Back in 2014, it might have been enough to know a few pandas functions and fire up a Jupyter notebook.
But in 2024, businesses have no appetite for proof-of-concept models stuck inside Jupyter notebooks. They don't need snazzy demos; they need full-stack solutions which deliver tangible value from messy, real-world data.
Pau Labarta Bajo
An ML model inside a Jupyter notebook has a business value of $0.
Unforuntately, many DS courses only focus on model-building and do not teach the full suite of skills required for modern data science.
The result is a mismatch between the skills graduates possess and the skills organisations need:
McKinsey & Company, 2024
We estimate that as much as 90 percent of the failures in ML development come not from developing poor models but from poor productization practices and the challenges of integrating the model with production data and business applications.
This course is different.
Our philosophy is that a data scientist shouldn’t just build models. Instead, they should be a "jack of all trades" actively involved in the entire lifecycle of an ML project.
The FSDS course has been built to reflect this philosophy. It consists of 6 modules designed by a team of professional data scientists, based on years of experience in industry. It'll give you all of the full-stack skills required of a modern data scientist, from data analysis to machine learning, A/B testing, MLOps and more.
By the end of the course, you'll be a highly effective programmer with a comprehensive set of data/MLOps skills and a sleek portfolio of projects to show prospective employers.
If that sounds up your street, we'd love to have you join us.
Go from the classroom to the real world
Real-world challenges
Tackle industry-relevant problems designed to hone your skills in Python, SQL, LLMs, MLOps, and more.
Hate speech classifier API
Use advanced techniques from NLP and deep learning to build and deploy a model.
Production-grade code
In the real world, data scientists don't just write production code in Jupyter notebooks.
Many other courses neglect this fact and don't teach tools like Git, GitHub and VS Code which you'll need in the real world.
Our projects are different.
We'll show you how to set up a holistic development environment and write code like a pro, ready to slot into a real-world data team.
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.metrics import precision, recall, roc_auc
import lightgbm as lgbm
def load_data(source: str):
df = pd.read_csv('data.csv')
df.set_index('id', inplace=True)
...
if __name__ == '__main__':
df = load_data('user_logs.csv')
...Showcase your portfolio
A model stuck on your local machine is no use to anyone.
Once you've built your solution, you'll productionize it and add a working demo to your portfolio - perfect for showing to prospective employers.
(We'll teach you how.)
joebloggs.com
Joe Blogg's Portfolio
Get feedback
In the real world, developers don't work in isolation. They work in teams.
After you've finished your project, push it to GitHub. This gives you the chance to practice a real-life, async Git workflow and get personalised feedback from a professional.
Minor correction at /api/predict/text
Hey @joe47, great job!
I just noticed one small issue at the endpoint /api/predict/text.
This blog post might help. Let me know if you have any questions :-)
One more important thought here.
Build 3 production-grade ML projects
Project 1
Price optimization with scikit-learn and XGBoost
Build a real-time, ML-powered mobile phone price prediction API.
🧑🏫 1 hour of video lessons
🧪 What you'll do
- ✅Webscraping - Build an automatic Ebay Deals crawler using Selenium, BeautifulSoup and cron job
- ✅Machine learning - Build a production-grade ML model with scikit-learn and XGBoost
- ✅MLOps - Deploy your model to a live website at yourportfolio.com
Project 2
Time series forecasting with LightGBM
Build a batch inference ML pipeline and deploy it to GitHub Actions.
🧑🏫 1 hour of video lessons
🧪 What you'll do
- ✅Build an end-to-end solution for house price prediction using machine learning algorithms and advanced feature engineering techniques
- ✅Build an automatic Ebay Deals crawler using Selenium, BeautifulSoup and cron job
- ✅Build a secure Chatbot with FastAPI and ChatGPT
Project 3
Supply chain optimization with OR-Tools
Covers the basics of MLOps, with a focus on practical skills in Linux, Python, Bash Scripting, WebAPI, Docker, Kubernetes, CI/CD, Cloud, Monitoring, and IaC.
🧑🏫 1 hour of video lessons
🧪 What you'll do
- ✅Build an end-to-end solution for house price prediction using machine learning algorithms and advanced feature engineering techniques
- ✅Build an automatic Ebay Deals crawler using Selenium, BeautifulSoup and cron job
- ✅Build a secure Chatbot with FastAPI and ChatGPT
Includes 3 bonus resources to get you job-ready (worth $190)
No course (including this one) can promise you a job. But one thing is for certain - to get a data science job, you'll need to showcase your skills.
This course is specifically designed with this objective in mind. By the end of the course, you'll have a suite of advanced, end-to-end data science projects to help you stand out to prospective employers.
Then, to help you showcase these projects, we'll give you a free portfolio template, resume template, and 30-min 1-2-1 career call.
Portfolio template
Get a free responsive website template at yourname.com, free to deploy on GitHub Pages or CloudFlare Pages. Normal price $23.
Resume template
Get a resume template for building your CV and cover letter.
121 career consultation
Some text will go here. And then a bit more as well, just to illustrate what the text wrapping should look like.
The wall of love
None of the testimonials on this page were paid for or coerced in any way.
If you want to feature here too, tweet me. Or else! 👊
Stephen Martin
This is hands down one of the best I’ve read in regards to beginning a career in Data Science. People trying to get into Data Science need to realize that there is a lot of work that goes into becoming a Data Scientists. You have to code and build daily. I like that. That worked for me.
Anna Zawadzka
I’m just trying to expand my portfolio, this is incredibly helpful! I also appreciate that you highlight the "find data for your problem" approach which promotes problem solving as opposed to "find problem for your data", which can lead to fancy but impractical results. Great job!
Wes Lee
Thanks Matt, I really like how you talk about putting a fresh spin on the tried and tested data sets and making it uniquely yours
Monicah Hinga
You've hit the nail on the head about the need to apply and adapt code to real-world problems, beyond the confines of course-specific IDEs. The importance of learning additional tools and problem-solving frameworks cannot be overstated. Thanks for sharing your personal journey and practical tips
Sree Pemira
Very inspiring & informative for a beginner like me aspiring to be a data scientist
About your instructor
It's me!
On holiday in the New Forest, England
Hi! I'm Matt.
From walking to reading, every skill takes time to acquire. Coding isn't any different. Time and effort is all you need and we're here to show you the way.
FSDS includes a bonus module (Module 0) which will teach you the basics of coding in Python.
If you're an experienced coder, feel free to skip this module.
One price. No subscription.
Lifetime access.
Lifetime access
$197
Lifetime access to the Full-Stack Data Science course and all associated materials.
$197 - Buy nowWhat you get:
- ✅Access to the full course
- ✅Portfolio template
- ✅Question bank (1000+ questions)