I built ganfs: A Python package that uses GANs to automate feature selection for

<!-- SC_OFF --><div class="md"><p>Disclosure: I built LabelSets (labelsets.ai). Sharing the technical approach behind how we score dataset quality.</p> <p>THE PROBLEM</p> <p>Most

<!-- SC_OFF --><div class="md"><p>We kept running into the same problem every time we rented a GPU to run Ollama + OpenWebUI or ComfyUI, we'd spend the first 45 minutes

<!-- SC_OFF --><div class="md"><p>Been heads down building graphify since April 5th. The idea: instead of stuffing your entire codebase into context every session, build a

<!-- SC_OFF --><div class="md"><p>Been a UFC fan for years and recently had some spare time to build a web app that uses Machine Learning to predict which fighter will win as a

<!-- SC_OFF --><div class="md"><p>Hey everyone,</p> <p>I’ve been building an end-to-end Air Quality (PM2.5) forecasting pipeline for 4 countries (US, UK, India, Australia) using

<!-- SC_OFF --><div class="md"><p>The modern ML (LLM) compiler stack is brutal. TVM is 500K+ lines of C++. PyTorch piles Dynamo, Inductor, and Triton on top of each other. I built

<!-- SC_OFF --><div class="md"><p>PyRadiomics is the de facto standard for radiomic feature extraction, but it's CPU-only and takes ~3 seconds per scan. At scale, that's a

<!-- SC_OFF --><div class="md"><p>We added native anomaly detection in Stratum, our columnar analytics engine for the JVM. Train and score isolation forest models entirely from