Object Types
Module Summary
Discover how Object Types define the nouns of your business — Employees, Flights, Transactions — and give raw rows real meaning.
Defining an Object Type
An Object Type maps a backing dataset to a concept in the real world. You give it a display name, choose a primary key column, and map columns to Properties with human-readable names and descriptions.
For example, you might back an "Employee" Object Type with a dataset whose columns are
emp_id, full_name, dept_code, and start_dt. The Ontology presents these as Employee ID, Full Name, Department, and Start Date.Properties and Primary Keys
Every Object Type has a primary key — the property that uniquely identifies each object instance. Choose carefully: it should be stable and never change (like an ID, not a name).
Properties are typed: string, integer, double, date, timestamp, boolean, or even arrays and structs. The types you choose affect how Workshop widgets render and how Functions query.
Search and Discoverability
Once an Object Type is published, it becomes searchable across the entire platform. Users can find objects through global search, Object Explorer, or within any Workshop app. This is the power of the Ontology — data is no longer locked inside a pipeline; it's a first-class citizen of the platform.
Key Takeaways
- An Object Type maps a backing dataset to a real-world concept.
- The primary key uniquely identifies each object instance.
- Properties are typed columns with human-readable names.
- Published Object Types are searchable across the entire platform.