The Data Vault Handbook - Concepts and Applications

3.3. QUERY ASSISTANT TABLES Finally, this section focuses on entities that serve a distinct role in enhanc- ing the performance and efficiency of querying in the Data Vault. Unlike the core entities, which primarily manage the raw data, special entities like Point-in-Time (PIT) and Bridge tables are composed of system generated data and reside within the Business Vault. Their primary function is to index the subset of data relevant for reporting, by doing that they simplify the joins necessary to retrieve information and accelerate querying. These entities are not part of the core data model but are essential for tuning performance and optimizing reporting workflows. 3.3.1. POINT-IN-TIME TABLES A PIT table addresses the complexity of querying data from multiple Sat- ellites connected to either a Hub or Link, especially when changes to the business objects in these Satellites occur at different times. In a Raw Data Vault, this leads to challenges in retrieving a consistent snapshot of a busi- ness entity that is loaded by multiple sources at a specific point in time, as updates across different systems do not happen simultaneously. PIT tables solve this problem by aligning timelines and therefore simplifying the re- trieval process, allowing for faster and more efficient queries without the need for complex time-based joins, ultimately improving performance. A PIT table functions as a generated index that tracks the specific state of each satellite associated with a Hub or a Link at a designated snapshot in time. Its granularity will depend on the requirements of the target locat- ed on the Information Mart layer. These tables are not typically exposed to business users but are integral in extracting data from the Raw Data Vault into the Information Mart layer, reducing the complexity of navigating mul- tiple satellite records over time.