Hopsworks

Hopsworks provides an AI Lakehouse platform designed to unify data science, data engineering, and machine learning. The company was co-founded in 2017 by CEO Jim Dowling, originating from academic research at KTH Royal Institute of Technology and the open-source community. Dowling, who has a PhD in distributed systems and AI, identified the need for a more integrated and scalable data platform for AI, leading to the creation of Hopsworks. The company was formally established with venture capital in 2018 when eleven people from the research environment started the company.

The core offering is a modular AI Lakehouse that includes an advanced, Python-centric feature store. This platform centralizes and governs data for AI models, enabling organizations to build and operate batch, real-time, and large language model (LLM) systems. It is designed to work with existing data ecosystems, supporting open table formats like Hudi, Delta, and Iceberg without requiring data migration. Key features include a high-performance query service, vector and similarity search for generative AI applications, feature monitoring, and robust data governance. Hopsworks supports popular frameworks such as Spark, TensorFlow, and PyTorch and can be deployed on-premises or on any major cloud provider like AWS, Azure, and GCP.

Hopsworks operates on a business model that includes open-source, self-managed commercial, and fully-managed cloud services. The platform is offered in a free community version and a commercial enterprise version with advanced features. The company generates revenue by selling its platform to enterprise clients, helping them to manage the entire MLOps lifecycle, from development to production. Its target clients are data science and engineering teams in various industries, with customers including American-First Credit Union, PaddyPower-Betfair, and the Swedish Public Employment Service.

Keywords: AI Lakehouse, feature store, MLOps, data engineering, machine learning platform, real-time AI, data governance, Python-centric, large language models, model deployment, data pipelines, vector database, batch processing, model monitoring, data versioning, sovereign AI, enterprise AI, generative AI, cloud-native data platform, ML infrastructure