Blog
Notes on building AI systems. Practical stuff, not thought leadership.
Fraud Detection: A Practical Approach
Building fraud detection systems is less about fancy algorithms and more about understanding how fra...
MLOps: What It Is and Why You Need It
Getting a model working is just the beginning. MLOps is the practice of keeping it working in produc...
Recommendation Engines: Beyond "Customers Also Bought"
Simple collaborative filtering is table stakes. Modern recommendation systems combine multiple signa...
AI for Customer Support: What's Real in 2024
Chatbots were going to replace support teams. That didn't happen. Here's what AI can actually do for...
Demand Forecasting for Retail
Better forecasts mean less stockouts and less overstock. Here's how to build forecasting systems tha...
When to Use (and Not Use) Deep Learning
Deep learning gets all the attention. But for many business problems, simpler approaches work better...
Building ML Teams: What to Look For
Hiring for ML is different from hiring for traditional software. Here's what we look for when buildi...