-
Transformers & Vision Transformers: A Deep Technical Guide
Deep-dive notes on attention, ViT, engineering trade-offs, and common pitfalls.
-
GANs for Image-to-Image Translation: An Engineering Perspective
What it actually takes to build a production GAN system — from architecture choices to training stability to deployment feedback loops.
-
Vectors vs. Pixels: Two Ways to Search for the Same Thing in Geometric Data
How the same spatial search problem can be solved with coordinate-based indexing or image-based morphology, and what the trade-offs teach us about algorithm design.
-
Semiconductor Data Is Not Like Other Data: A Practical Guide for ML and Data Engineers
What makes manufacturing metrology data unique, how it differs from the datasets most ML practitioners are used to, and what to do about it.
-
Beyond Rendering: Rasterization, Topology, and the Bridge to Search
Why rasterization in engineering systems is a data transformation—not a drawing exercise—and how topology preservation enables reliable downstream search.