PinnedPublished inTDS ArchiveDocument Extraction is GenAI’s Killer AppThe future is here and you don’t get killer robots. You get great automation for tedious office workAug 13, 202410Aug 13, 202410
Vector Similarity Search is HopelessA pre-baked notion of similarity is inherently flawed and doomed to fail. We can do better.Dec 29, 202412Dec 29, 202412
Published inThe StartupBuilding an AI SaaS that makes money in under 24 hoursNot a get-rich-quick guide. But a “try stuff quick” guideNov 5, 20244Nov 5, 20244
Published inTDS ArchiveWhy Explainability Matters in AINot because we’re curious. Because we need to get shit done.Oct 20, 20243Oct 20, 20243
JPG format sucks. We can do something about itWe all use JPG photos on our websites and apps. In this post I’ll show how you can reduce image sizes by an additional 20–50% with a single…Jul 10, 2022Jul 10, 2022
Published inTDS ArchiveFrom the Gym to a Jupyter Notebook — Building a Squats Counter App in a DayJun 2, 20211Jun 2, 20211
Stripe is Choosing to become a Predatory LenderSo this is a quick note about something interesting I’ve come across. I own a Machine Learning Software-as-a-Service called Product Pix…Nov 16, 20201Nov 16, 20201
How I built my first Machine Learning Software-As-A-ServiceAnd guess what? It didn’t need any funding. But it did need human (and still does)Feb 23, 201914Feb 23, 201914
Published inTDS ArchiveWhat To Optimize for? Loss Function Cheat SheetSome tips in finding the right target for optimization, and how to figure it out for your use caseFeb 15, 20191Feb 15, 20191
Published inTDS ArchiveHow to (quickly) Build a Tensorflow Training PipelineHow to build an efficient training and pipeline in Tensorflow without getting lost in the woodsOct 26, 20186Oct 26, 20186
