h2o.ai

Founded in 2012, h2o.ai builds software that combines generative and predictive artificial intelligence for enterprises and public sector agencies. The company's core offering is a platform on which organisations can develop purpose-built GenAI applications running on their own private data, with deployment options that include airgapped, on-premises and cloud VPC environments.

The technical work spans agentic AI, generative and predictive modelling, automated machine learning, large language models, distributed machine learning, automatic feature engineering and model acceleration. Product lines include H2O AI Cloud, an end-to-end GenAI platform; H2O Driverless AI, for automatic feature engineering and model acceleration; H2O LLM Studio, for no-code training and tuning of large language models; h2oGPTe, enterprise AI agents with multi-model support; and H2O-3, a fully open-source distributed machine learning platform for Python, R and Spark.

More than 20,000 organisations worldwide use h2o.ai's open-source technology, and the company reports that over half of the Fortune 500 are customers. Named users include AT&T, Commonwealth Bank of Australia, Chipotle, Workday, Progressive Insurance and the NIH, spanning telecommunications, banking, insurance, food service, enterprise software, government and health. Technology and consulting partners include NVIDIA, Dell Technologies, Deloitte, Ernst & Young, Snowflake, AWS and Google Cloud Platform.

Engineering culture is shaped by an open-source ethos, epitomised by H2O-3, and by an emphasis on data privacy, security and customer control over AI infrastructure.