Writing

The quality fixes that took a working MCP server from 64 to 100/100 on Smithery: tool descriptions, dot notation naming, annotations, caching, smithery.yaml, and the skill that automates the checks.

Seven open-source PDF parsers tested on complaint forms: scanned pages, handwriting, AcroForm fields, and degraded fax. Where each parser breaks, and which one I would use in production.

How I built a wine price comparison tool for Singapore. Amex Platinum gives you S$200 back on wine, but are the base prices actually competitive? I answered that with Selenium scraping, bundle-aware comparisons, Brave-assisted Vivino resolution, and a guarded daily refresh pipeline.

How I built a Singapore job search and resume product with nightly crawls, RAG matching, a 7-stage tailoring pipeline, and 5 validation gates to keep edits factual.

Markets have moods. A Hidden Markov Model reads the volatility regime (calm, normal, or stormy) so you can trade with the weather, not against it. Interactive walkthrough with live visualizations.

An interactive walkthrough of Direct Preference Optimization: how to align LLMs with human preferences using a classification loss. No reward model, no RL, no PPO.

19 engineers, 19 A100 GPUs, two hours of Locust fire. I tuned 40+ vLLM configs and scraped peer metrics. Then someone with a 30-line cache proxy beat all of us.

Prefill vs decode. Why memory bandwidth matters more than FLOPS. VRAM budget math. Quantization methods (GPTQ, AWQ, FP8). Batching strategies. The theory behind every LLM serving decision.

An interactive walkthrough of the WMSS paper, from SFT saturation to logit mixing and gradient amplification. Use the sliders to see how weak agents can improve stronger agents.

MediaPipe face mesh, rembg segmentation, 25+ compliance checks, and country-specific rules for 6 countries. Here's how Photo ID Studio decides if your face is passport-worthy.

A deep dive into Trader Koo – from using a pre-trained YOLOv8 model to detect chart patterns, to solving the coordinate mapping problem, to running a five-layer detection ensemble every night.

Building LionWeather: tropical weather data from 2016 to 2024, LightGBM rainfall forecasting, SHAP explainability, animated radar, and a whole lot of lessons about Singapore weather.