A data lakehouse enables both batch and streaming analytics, distinguishing it from traditional data warehouses. Key features include:

• Transaction support: Ensures consistency with ACID transactions for concurrent reads/writes.
• Schema enforcement and governance: Supports schema evolution and integrity, with robust governance.
• BI support: Allows BI tools to use source data directly, reducing latency and costs.
• Decoupled storage and compute: Scales efficiently for more users and larger data sizes.
• Openness: Uses open storage formats like Parquet, accessible by various tools.
• Diverse data types and workloads: Supports everything from unstructured data to machine learning.
• End-to-end streaming: Facilitates real-time data applications without separate systems.