Databricks is a unified data analytics and artificial intelligence platform built around the lakehouse architecture, which combines the capabilities of data lakes and data warehouses into a single platform for data engineering, data science, machine learning, and business analytics. Founded in 2013 by the original creators of Apache Spark at UC Berkeley, including Ali Ghodsi, Matei Zaharia, and five other co-founders, Databricks is headquartered in San Francisco, California. The platform is built on and extends Apache Spark, providing a managed cloud environment for processing massive datasets and building AI applications. Databricks offers several integrated components. The Unity Catalog provides unified data governance across all data and AI assets. Delta Lake, an open-source storage layer, provides ACID transactions, schema enforcement, and time travel for data lakes. MLflow, another Databricks-originated open-source project, provides experiment tracking, model registry, model serving, and ML lifecycle management. Databricks SQL enables SQL analytics and dashboarding directly on lakehouse data. The platform includes Mosaic AI, its suite of AI and machine learning tools that encompasses model training, fine-tuning, serving, and monitoring. Mosaic AI Agent Framework supports building compound AI systems and retrieval-augmented generation applications. Databricks also offers Foundation Model APIs for accessing popular large language models and Vector Search for similarity search on embeddings. The platform runs on all major cloud providers including AWS, Azure, and Google Cloud, with customers deploying within their own cloud accounts for data security and compliance. Databricks pricing follows a consumption-based model using Databricks Units (DBUs), with rates varying by workload type and compute tier. The platform serves organizations of all sizes, from startups to the largest enterprises in the world, across industries including financial services, healthcare, retail, media, and technology.
databricks.com →