1 - About this Course

The Foundry Ontology is a “digital twin” of entities, events, and processes that are important to an organization. This semantic object layer provides a consistent API that you and your colleagues can use to build operational and interoperable applications. In the Data Lineage image below, we see the [Example Data] Flight object type in the Foundry Ontology is used in at least two operational applications (labeled as “modules” in Data Lineage) and links to a few other object types, such as aircraft and flight alerts.

The graph also illustrates that the [Example Data] Flight object type is “backed” by a dataset, which is the output of a scheduled build. This flow illustrates the central theme in this tutorial: data engineers must work closely with their app builder colleagues to craft pipelines and datasets that form the basis for Foundry Ontology objects and links.

⚠️ Course prerequisites

• DATAENG 07: Configuring Data Expectations: If you have not completed the previous course in this track, do so now.
• Necessary permissions to edit your organization’s Ontology. Contact you program administrator or Palantir Support if you need authorization.

Outcomes

Data pipelines are not isolated artifacts; they’re always built for some purpose. In this training, the purpose of our pipeline is to prepare datasets to back a set of linked Ontology object types. You’ll have hands-on practice configuring the Ontology and updating a pipeline to optimize the backing datasets and make them resilient to upstream data changes that would alter the shape of your Ontology. Finally, a collection of exercises will walk you through common troubleshooting steps.

Learning Objectives

1. Understand core Ontology concepts, basic use cases, and design decisions.
2. Search and filter object types in the Object Explorer.
3. Know how to create new object and link types in the Ontology Manager.
4. Optimize backing datasets for use in the Ontology in accordance with best practices.
5. Apply checks to your pipeline to ensure object types are resilient to upstream data changes.
6. Understand the Ontology storage and retrieval architecture.
7. Develop an experience-based understanding of "writeback" and its place in a data pipeline.
8. Practice troubleshooting common Ontology synchronization issues.

💪 Foundry Skills

• Use the Object Explorer to search, filter, and generally understand the Ontology.
• Use the Ontology Manager to create and edit ontology object and link types.
• Update pipeline transforms to optimize a backing dataset for synchronization with the Ontology backend.
• Create "writeback" datasets and position them in your data pipeline.
• Accommodate additive and destructive schema changes via the object storage interface in the Ontology Management Application.