Lakeflow Designer updates: Guardrails, multi-input joins and Genie Code
The September release of Lakeflow Designer brings rule validation with Guardrails, multi-input joins, operator creation with Genie Code, and more.
O que mudou
Use cases
Imagine you need to build visual data preparation workflows and now have new features to validate data quality, join multiple sources, and automate tasks. Lakeflow Designer expands the capabilities of the visual canvas.
- Validate data against rules for nulls, unique values, and accepted values using the Guardrails operator.
- Combine three or more tables in a single Join operator.
- Create custom user-defined operators using natural language descriptions with Genie Code.
- Import multiple files at once as source operators.
How to use
To create a new visual data prep workflow in Databricks with Unity Catalog enabled:
- Click New in the sidebar and select Visual data prep.
- Add source operators by dragging files or connecting tables.
- Configure new operators like Guardrails or Join directly in the side pane.
- Use Genie Code in the sidebar to generate Python functions and configurations from text descriptions.
Limitations and requirements
- Requires a Databricks workspace with Unity Catalog enabled.
- Requires
CAN USEpermission on at least one general-purpose compute resource (serverless or all-purpose).
- lakeflow designer
- unity catalog
- pipelines
- data prep
- genie code
Lakeflow Designer September Release
Resumo e explicação gerados por IA a partir da release note oficial. Em caso de dúvida, a documentação oficial da plataforma prevalece.
Relacionadas
- DatabricksGA
Repositórios de pacotes em nível de workspace para Databricks Apps
Administradores de workspace podem configurar repositórios padrão de pacotes Python e npm para Databricks Apps.
databricks apps · workspace settings - DatabricksGA
Redimensione partições de stateful streaming sem perder o checkpoint no Databricks
Agora é possível alterar o número de partições de shuffle em queries Structured Streaming e pipelines Lakeflow sem descartar o estado do checkpoint.
structured streaming · lakeflow - DatabricksAtualização
Google Gemini 4 Argon disponível no Unity Gateway do Databricks
O modelo multimodal Google Gemini 4 Argon já pode ser acessado via Foundation Model APIs e Unity Gateway.
unity catalog · foundation model apis