Machine learning tutorials with reproducible code
Kudos AI features original machine learning tutorials covering NLP, computer vision, time-series forecasting, and responsible AI. Every article includes the math, the intuition, and runnable code you can test immediately.
All notebooks are hosted on GitHub with one-click Google Colab launchers, so you can reproduce every experiment without any setup.
Latest articles
Long-form guides that blend theory with practice. Each one includes math breakdowns, Python code, and links to the full notebook.
February 21, 2026
A complete guide to RAG: why pure LLMs fail, how retrieval + generation work together, embeddings, vector search, chunking strategies, and production architecture.
Read the tutorialSeptember 28, 2025
Support Vector Machines (SVMs) are one of those algorithms every data scientist hears about early on. Once you understand "maximize the margin" and the idea...
Read the tutorialSeptember 14, 2025
Choosing the right algorithm is half the battle in machine learning. This deep dive explains the Top 10 ML algorithms with plain-English intuition and Python snippets.
Read the tutorialWhat you'll find here
In-depth tutorials on NLP, computer vision, forecasting, and responsible AI with diagrams and runnable code.
Read the articlesHands-on builds with datasets, architectures, and Colab launchers you can run immediately.
See the projectsNeed help with a machine learning project? Let's discuss how we can support your team.
Get in touchFeatured projects
Each project links to a GitHub repository and Google Colab notebook you can run immediately.
Q-learning Tic-Tac-Toe bot and reward-tuning walkthroughs.
Open notebooksSRGAN super-resolution and OpenCV face detection.
Explore buildsS&P 500 models and demographic forecasts with ARIMA/LSTM.
View forecastsMultilingual T5 summarizer and Wikipedia-trained chatbot.
View NLP projectsAbout this site
Kudos AI is an independent educational platform created by Anas HAMOUTNI, a statistical engineer with a background in actuarial science, quantitative finance, and applied mathematics. The site is built around one idea: the best way to understand AI is to combine rigorous theory with hands-on practice.
Every tutorial on this site starts from the mathematical foundations, loss functions, gradient updates, attention mechanisms and walks through them step by step before presenting working Python code. All notebooks are open-source on GitHub, and most include a one-click Google Colab launcher so you can run every experiment yourself without installing anything.
The blog covers a wide range of topics across natural language processing, computer vision, time-series forecasting, and responsible AI. Articles range from introductory guides (like gradient descent explained from scratch) to deep technical breakdowns of architectures like Transformers and training techniques like RLHF.
Beyond tutorials, Kudos AI hosts a growing collection of reproducible ML projects, from SRGAN image super-resolution to Q-learning game agents and a set of interactive browser games that teach classification, optimization, and linear algebra concepts through play. Whether you are a student, a self-taught practitioner, or a professional looking to deepen your understanding, you will find original, carefully written content here.