Contextual search is driven by two platform-native components:
1. Zing Search Engine — ServiceNow's built-in text indexing engine that computes document scores based on frequency, sequence, and weight of search terms, split across multiple database shards for parallel querying. ServiceNow When a user types in a field, Zing performs a full-text search and returns relevancy-scored results in real time.
2. Declarative CXS Configuration Tables — A set of cxs_* tables (cxs_search_context, cxs_table_config, cxs_rp_config, resource configs, filter configs) that define where results appear, what sources to search (KB, Catalog, etc.), and how to filter them — all without writing code.
Optional: Predictive Intelligence — With the Predictive Intelligence for Contextual Search plugin installed, an ML-based similarity resource is added alongside Zing, enabling semantic matching beyond just keywords.
The only place custom script enters the picture is through Scripted Filter Configurations, which are optional and used only when you need dynamic, user-specific filtering logic. The core search mechanism is entirely platform-built.
