This document discusses H2O.ai, an open source in-memory machine learning platform. It can perform distributed machine learning on large datasets using algorithms like generalized linear modeling, gradient boosted machines, random forests, and deep learning. The platform provides APIs and interfaces for R, Python, Scala, Spark, and other languages. It can handle big data from sources like HDFS, S3, and NFS without sampling. The document includes an overview of H2O's architecture and demonstrates its use on a bike sharing dataset with over 10 million rows.