Bring data producers and consumers together to define data contracts in a facilitated workshop.
A defined and agreed upon data contract between data producers and consumers.
- Facilitator
- Neutral moderator and typist
- Should know the Data Contract Specification and its tools well
- Get the authors of the Data Contract Specification as facilitators for your workshop.
- Data producer
- Product Owner
- Software Engineers
- Data consumers
- Product Owner
- Data Engineers / Scientist / Analyst
Recommendation: keep the group small (not more than 5 people)
- Show data contract the whole workshop on the screen (projector, screenshare, ...)
- Facilitator is the typist
- Facilitator is moderator
- Data Producer and Data Consumers discuss and give commands to the facilitator
- Info (get the context)
- Examples (example-driven facilitation)
- Model (you will spend most of your time here)
- Use the Data Contract CLI to test the model against the previously created examples:\
datacontract test --examples datacontract.yaml
- Use the Data Contract CLI to test the model against the previously created examples:\
- Quality
- Terms
- Servers (if already applicable)
- Start with a "local" server with actual, real data you downloaded
- Use the Data Contract CLI to test the model against the actual data on a specific server:\
datacontract test datacontract.yaml - Switch to the actual remote server, if applicable
- Open the starter template in the Data Contract Editor and get going. If you lack an experienced facilitator, ignore any validation errors and warnings within the editor.
- Use the Data Contract Editor to share the results of the workshop afterward with the participants and other stakeholders.
- Use the Data Contract CLI to validate the data contract after the workshop.
- This data contract workshop could be a followup to a data product design workshop using the Data Product Canvas, making the offered contract at the output port of the designed data product more concrete.