This course gives you a practical introduction to tabular foundation models, the technology bringing foundation models to the rows and columns that run most organizations. You will learn what tabular data is, why traditional machine learning workflows can feel heavy, and how a model pre-trained across many datasets can generalize to a spreadsheet it has never seen. We will cover classification and regression, filling in missing values from patterns already present in your data, and the shift from building models to simply asking what you want to predict.

The second half moves from concept to practice with TabH2O, H2O.ai's foundation model for tabular data. You will install it as an Excel add-in, load a dataset, select a target column, choose the task type, and generate predictions with a single click. If you have never built a model before, you will see how far you can get from a spreadsheet alone. If you already work in machine learning, you will see where a foundation model speeds up early exploration, and where a traditional workflow is still the better choice. Data privacy, real-world use cases across industries, and agentic AI integration are covered along the way. No prior machine learning or coding experience is required. TabH2O is free to use, and the sample dataset and setup links used in the demo are included with each lesson, so you can follow along in your own spreadsheet and try it on your own data straight afterwards.