A practical approach using Google Cloud — Matrix multiplications are still widely used for Recommendation Systems. And they are a great starting point if you want to build a quick and easy solution to deliver great recommendations to your customers. Nothing wrong with that. Keep it simple until you make it complicated. In this article, we want…

Deploying and serving any kind of machine learning model at any scale. — Companies frequently deploy their models to virtual machines (Google Compute Engine or even on-prem machines). This is something that should be avoided. Google Cloud provides a dedicated service called Vertex AI Endpoints to deploy your models. Vertex AI Endpoints provides great flexibility compared with easy usage. You can keep it…

How to build a fully automated recommendations pipeline in under 45 minutes — Matrix factorizations are widely used for Recommendation Systems. They are a great and easy starting point if you want to build a quick and easy solution to deliver great recommendations to your customers. This article covers the steps needed to automate a BigQueryML Matrix Factorization recommendation pipeline. Including automatic handling…

Since Open AI’s DALL·E 2 and Google’s Imagen, AI art and generally creating images with machine learning models got immense attention. Basically, everyone wants to create AI Images. And with the tools, available anyone can do it. This article is a get-started guide on how you can create your own…

A simple and scalable approach with custom container — I regularly guide customers on how they can scale their training on the Cloud. While every customer has unique requirements and expectations, the overall process is the same. In this article, we train a transformer model (DistilBert) with Vertex AI. As a machine learning Framework, I use TensorFlow. It will…

Reducing your ML inference costs can be so easy. 4 steps I regularly recommend to companies. — In this article, I cover the most common ways of saving costs. Some of them are just applying common sense, some are a nice way to utilize another service, while others require more technical knowledge. Cloud Run instead of Vertex AI Endpoints One of the major disadvantages of Vertex AI Endpoints is the fact that there is…

A personal opinion based on workshops with many different customers — Back in the days, at a time when ML was not all over the place. Companies already did a lot of data processing for other purposes. A large number of Cloud Composer workflows got created. Nowadays companies often try to implement machine learning workflows into their existing Cloud Composer workflows…

How to interpret and understand ML models with explainable AI (XAI) — Explainable AI is as the name already suggests it’s all about explaining predictions our models are making. It helps you to understand why your machine learning model is making certain decisions. Useful to prove if the model is focusing on the right patterns or to find issues like bias or…

The good and the bad stuff, no seriously just a deep dive into the product — Putting a similarity index into production at scale is a pretty hard challenge. It requires a whole bunch of infrastructure working closely together. You need to handle a large amount of data at low latency. …