The Two Numbers That Predict AI Agent ReliabilityBack then, I felt lost while writing automated Playwright tests. I'd turned to Claude Code to help speed up the process, but I kept hitting the same wall. The generated tests were redundant, surface-lApr 25, 2026路10 min read
From "It Works" to "Why It Works": A Call for Deeper Understanding in Data ScienceOct 5, 2025路4 min read
Building Intuition for Convolutional Neural NetworksThe motivation behind Convolutional Neural Networks (CNNs) comes from the inability of traditional dense neural networks to perform well on image classification tasks. Why is that? A dense network, also known as a fully-connected network, treats an i...Aug 24, 2025路8 min read
Building a Neural Network from Scratch in Python and NumPyBridging the Gap Between the Formulas and the Feeling of 'Learning'Aug 8, 2025路11 min read
Why ML Fails at Stock/Crypto PredictionThe use of ML (machine learning) has been on the rise for the past 10 years. The idea of machine learning has been a trend for quite a while now and many people coming from different domains try to grasp the topic in hopes for a better opportunity ca...Jun 20, 2025路3 min read
Reinforcement Learning is the inevitableThe internet is full of bad data, and that is where the training data for LLMs are coming from. Some are polluting the internet with bad data out of spite. Most just spam the internet of AI generated data for profit One day, we might see a future whe...Jun 18, 2025路2 min read
PCA as a Last ResortWhy You Should Do Feature Engineering and Feature Removal Before Immediately Using PCA.Apr 28, 2025路3 min read
Chunking Methods for RAG: What and WhyThe Day My RAG System Failed Meet Charlie, a developer who learned a valuable lesson about RAG systems the hard way. (Let鈥檚 just call him Charlie but we all really know who he is. 馃槈) Last year, Charlie built what he thought was the perfect RAG (Retr...Mar 3, 2025路10 min read
