Library

Hey, Rahul here! 👋 Each week, I publish long-form ML+AI posts covering ML, AI, and System design for MLwhiz. Paid subscribers also get how-to guides with full code walkthroughs. I publish occasional extra articles. If you’d like to become a paid subscriber, here’s a button for that:

People ask me this all the time: “Where’s your post on LinkedIn?”

Most of the time, I am digging through my own archive on a phone and sending links one at a time. It’s not great. So I sat down, tagged every post I’ve written, organized them by theme, and put them all in one place. This is that page.

If you’re looking for something specific (your first ML job, learning PyTorch, building a recommender, shipping GenAI to production), start with the right category below.

The library is roughly organized as: a starter shelf, then series-driven deep dives, then thematic buckets for the standalone posts.

If you can’t find what you need, reply to any post and let me know. I’ll either point you to something or add it to the writing queue.

Jump to a section:

• Start here

• RecSys for MLEs

• GenAI, LLMs, RAG

• Transformers, BERT, NLP

• Computer vision and deep learning

• Classical ML algorithms

• Statistics and probability

• Python for ML engineers

• Pandas, Spark, scaling data

• MLOps, deployment, production

• Coding interview algorithms

• Career, interviews, and work

• Tools and setup

• The MLWhiz newsletter

• Asides and opinions

---

Start here

A handful of evergreen entry points. If you’re new to the blog, read these first.

• How did I learn Machine Learning?

• 6 Important Steps to build a Machine Learning System

• The Real-World Guide to Machine Learning Careers in 2025

• Become a Machine Learning Engineer with these 10 resources

• A Day in the Life of an ML Engineer

---

RecSys for MLEs

The current series. From “how does a recommender actually work” to “how does Meta serve a trillion-parameter HSTU in production.” Each post stands alone, but they read best in order.

1. RecSys Fundamentals: The Art and Science of Digital Matchmaking

2. How Recommendation Systems Learned to Think

3. The 3-Stage Funnel Behind Every Modern Recommender System

4. How YouTube Finds Your Next Video in Milliseconds

5. From Candidates to Clicks: The Engineering Anatomy of Ranking

6. 3 Modern Approaches to Solving Cold Start in RecSys

7. Your Ranking Model Is Right. Your Recommendations Are Wrong (re-ranking)

8. From RNNs to Transformers: Building Sequential Recommenders (Part 9a)

9. From Random IDs to Semantic IDs: Building a Generative Recommender from Scratch (Part 9b)

10. HSTU: How Meta Built a Trillion-Parameter Recommender That Actually Scales (Part 9c)

Related: Vector Search at Scale: The Production Engineer’s Guide

---

GenAI, LLMs, and RAG

Everything LLM-shaped. The architectural journey, prompt engineering, fine-tuning, RAG (basic and agentic), production reality checks.

• GenAI 101: How Generative AI is Rewriting the Rules for ML Engineers

• A Review of the Architectural Journey of LLMs: Key Milestones from 2017 to Present

• Everything you need to learn about GPT: How Does ChatGPT Work?

• Everything Programmers need to learn about GPT: Using OpenAI

• The Art and Science of Prompt Engineering: A Comprehensive Guide

• Prompt Engineering: Why Your Tech Career Shouldn’t Rest on Vibes

• GenAI Series: My Tryst with AI-Assisted “Vibe Coding”

• Building RAG Applications From Scratch: The No-BS Guide

• Building Production-Grade Agentic RAG: Intelligent Recommendation Engines

• What is GraphRAG and how it works

• Fine-Tuning LLMs: Your Guide to PEFT, QDoRA, and Other Nifty Tricks

• The #1 Reason Your GenAI Project Will Fail in Production

• What Production Deployments Taught Me About ReAct vs Function Calling

• Did Apple Just Burst the AI Reasoning Bubble? I don’t think so

• Which AI Subscription is Actually Worth Your Money?

---

Transformers, BERT, and NLP

The NLP Learning Series is a complete 4-part walkthrough from preprocessing to transfer learning. Plus standalone deep dives on Transformers, BERT, and word embeddings.

NLP Learning Series (4 parts, text classification end to end):

1. Text Preprocessing Methods for Deep Learning

2. Conventional Methods for Text Classification

3. Attention, CNN and what not for Text Classification

4. Transfer Learning Intuition for Text Classification

Transformers and BERT:

• Understanding Transformers, the Data Science Way

• Understanding Transformers, the Programming Way

• Understanding BERT with Huggingface

• Explaining BERT Simply Using Sketches

Other NLP:

• What are Word2Vec Embeddings?

• What Kagglers are using for Text Classification

• Chatbots aren’t as difficult to make as You Think

• Using Deep Learning for End to End Multiclass Text Classification

• Adding Interpretability to Multiclass Text Classification models

• What my first Silver Medal taught me about Text Classification and Kaggle

---

Computer vision and Deep learning

Object detection, instance segmentation, GANs, image classification, and the PyTorch tooling that surrounds them.

Object detection and segmentation:

• Object Detection: An End to End Theoretical Perspective

• Demystifying Object Detection and Instance Segmentation for Data Scientists

• Implementing Object Detection and Instance Segmentation for Data Scientists

• How to Create an End to End Object Detector using Yolov5

• How I Created a Dataset for Instance Segmentation from Scratch

GANs:

• An End-to-End Introduction to GANs using Keras

• A Layman’s Introduction to GANs for Data Scientists using PyTorch

PyTorch and image classification:

• The Most Complete Guide to PyTorch for Data Scientists

• A layman's guide to moving from Keras to Pytorch

• End-to-End Pipeline for Multiclass Image Classification

• Pretrained Neural Networks: What are they?

---

Classical ML algorithms

The foundation. Feature work, evaluation, hyperparameter tuning, decision trees, XGBoost, time series, imbalanced data, interpretability.

• Machine Learning Algorithms for Data Scientists

• A Generative Approach to Classification

• The 5 Classification Evaluation metrics every Data Scientist must know

• A Layman’s guide to ROC Curves And AUC

• The 5 most useful Techniques to Handle Imbalanced datasets

• The 5 Sampling Algorithms every Data Scientist need to know

• The 5 Feature Selection Algorithms every Data Scientist should know

• The Hitchhiker’s Guide to Feature Extraction

• The Simple Math behind 3 Decision Tree Splitting criterions

• How to find Feature importances for BlackBox Models?

• Bayesian Bandits explained simply

• Using XGBoost for time series prediction tasks

• Using Gradient Boosting for Time Series prediction tasks

• Good Feature Building Techniques: Tricks for Kaggle

• Exploring Vowpal Wabbit with the Avazu Clickthrough Prediction Challenge

• Hyperopt: A Bayesian Parameter Tuning Framework

• Automate Hyperparameter Tuning for your models

• 100x faster Hyperparameter Search Framework with Pyspark

• Lightning Fast XGBoost on Multiple GPUs

• How to run your ML model Predictions 50 times faster?

---

Statistics and probability

Distributions, p-values, confidence intervals, MCMC, and a few “your intuition is lying to you” pieces.

• P-value Explained Simply for Data Scientists

• Confidence Intervals Explained Simply for Data Scientists

• The story of every distribution: Discrete Distributions

• My Tryst With MCMC Algorithms

• Behold the power of MCMC

• Maths Beats Intuition probably every damn time

• Five Cognitive Biases In Data Science (And how to avoid them)

---

Python for ML engineer

Pandas, idioms, generators, decorators, async, dunders, regex, OOP. The Python-language layer of the job.

Idioms and patterns:

• Object-Oriented Programming Explained Simply for Data Scientists

• A primer on *args, **kwargs, decorators for Data Scientists

• Do you know about the Magic Methods in Python?

• Use Iterators, Generators, and Generator Expressions

• Start using Python defaultdict and Counter

• Minimize for loop usage in Python

• How and Why to use f strings in Python3?

• The Ultimate Guide to using the Python regex module

• 3 Programming concepts for Data Scientists

Concurrency and performance:

• The Primer on Asyncio that I Wish I Had

• How Can Data Scientists Use Parallel Processing?

• Make your own Super Pandas using Multiproc

Visualization:

• 3 Awesome Visualization Techniques for every dataset

• Create basic graph visualizations with SeaBorn

• Plotly Express: Python’s One Liner graph creation library with animations

---

Pandas, Spark, and scaling data

Everything in the “make the data move faster” bucket. Pandas tricks, GPU pandas, PySpark from basics to production tips.

Pandas:

• Minimal Pandas Subset for Data Scientists

• Pandas For All: Some Basic Pandas Functions

• Add this single word to make your Pandas Apply faster

• Minimal Pandas Subset for Data Scientists on GPU

• Bamboolib: Learn and use Pandas without Coding

• How to use SQL with Pandas?

Spark and big data:

• The Hitchhikers guide to handle Big Data using Spark

• The Most Complete Guide to pySpark DataFrames

• Learning Spark using Python: Basics and Applications

• 5 Ways to add a new column in a PySpark Dataframe

• Practical Spark Tips for Data Scientists

• Accelerating Spark 3.0 with NVIDIA GPUs on Google Dataproc

• Hadoop, Mapreduce and More: Part 1

• Hadoop Mapreduce Streaming Tricks and Techniques

• Working with Large-Scale One-Hot Encoding

Graph algorithms on big data:

• 4 Graph Algorithms on Steroids for data Scientists with cuGraph

---

MLOps, deployment, and production

Flask, FastAPI, Streamlit, Docker, EC2. Taking a model from notebook to a URL someone else can hit.

• Take your Machine Learning Models to Production with these 5 simple steps

• Deploying ML Apps using Python and Flask

• How to write Web apps using simple Python for Data Scientists? (Streamlit intro)

• How to Deploy a Streamlit App using an Amazon Free ec2 instance

• Share your Projects even more easily with this New Streamlit Feature

• A Layman’s Guide for Data Scientists to create APIs in minutes (FastAPI)

• Deployment could be easy: A Data Scientist’s Guide to deploy an Image detection model

• The 5 Principles That Turn ML Projects Into Business Impact

---

Coding interview algorithms

The patterns that come up in ML engineer interviews. Binary search, DP, linked lists, trees, graphs.

• Solve almost every Binary Search Problem

• Dynamic Programming for Data Scientists

• A simple introduction to Linked Lists for Data Scientists

• Handling Trees in Data Science Algorithmic Interview

• The one Graph Algorithm you need to know (connected components)

• The 5 Graph Algorithms that you should know

---

Career, interviews, and work

Interview prep, system design, and advice from someone who’s been on both sides of the hiring table.

Interview prep:

• How to Crack FAANG Coding Interviews?

• Crack ML System Design Interviews Like a Pro: Part 1

• Crack ML System Design Interviews Like a Pro: Part 2

• How I cracked my MLE interview at Meta

• 3 Mistakes you should not make in a Data Science Interview

• 5 tips for getting your first Data Science job

Career direction:

• The Real-World Guide to Machine Learning Careers in 2025

• Become a Machine Learning Engineer with these 10 resources

• Become an ML Engineer with these courses from Amazon and Google

• A Day in the Life of an ML Engineer

• Top advice for a Data Scientist

• 5 Essential Business-Oriented Critical Thinking Skills For Data Scientists

• How should I start Reading code?

• How to Look like a 10x developer

• The Action-Planning Paradox: Balancing Execution and Strategy

---

Tools and setup

Workstations, IDEs, AI assistants, and dev environment.

AI assistants and Claude Code:

• I Use Claude Code Every Day. Here’s the Setup That Actually Matters

• Supercharge Your Claude: Adding Internet Powers with Brave Search

IDEs and editors:

• Create an Awesome Development Setup for Data Science using Atom

• Why Sublime Text for Data Science is Hotter than Jennifer Lawrence

• 3 Great Additions for your Jupyter Notebooks

Workstations and infra:

• Stop Worrying and Create your Deep Learning Server in 30 minutes

• A definitive guide for Setting up a Deep Learning Workstation with Ubuntu

• Creating my First Deep Learning + Data Science Workstation

Shell, SQL, and basics:

• Shell Basics every Data Scientist Should know: Part I

• Shell Basics every Data Scientist Should know: Part II (AWK)

• Learning SQL the Hard Way

• How to Set Environment Variables in Linux?

• Data Science 101: Playing with Scraping in Python

---

The MLWhiz Weekly newsletter

Curated AI/ML/RecSys/GenAI links each week. New issues every Sunday.

• Issue #1

• Issue #2

• Issue #3

• Issue #4

• Issue #5

• Issue #6

---

Asides and opinions

Off-topic essays, opinion pieces, and the occasional rant. Read these when you want a break from the technical posts.

• Why LLMs Won’t Replace Programmers: A Reality Check

• Claude Code vs. Your ML Career: A 2026 Reality Check

• Don’t Democratize Data Science

• Tech Professionals in UK: Personal Finance Notes

• A Newspaper for COVID-19: The CoronaTimes

---

Closing

That’s everything. ~160 posts across 12 years of writing.

If you got value from a specific post, share it with someone who’d benefit. If a topic is missing, reply to any of my posts and tell me what’s on your wish list. I read every reply.

If you want to support the work, the easiest way is to become a paid subscriber. Paid subscribers fund the deep-technical posts, which take real time to research, code, and illustrate.

Thanks for reading.