Databricks is evolving its Partner Well-Architected Framework (PWAF) to meet the rapid pace of AI development and platform updates. According to the announcement, the updated framework provides AI-ready architecture guidance, technical standards, and best practices for partners building on, connecting to, or sharing data through Databricks.
The PWAF now spans all three core partner architectures: Built-On, Connected, and Data Collaboration. This comprehensive approach is designed to accelerate development and align with platform best practices as partners increasingly build data and AI applications.
What's New in the Partner Well-Architected Framework
Since its February launch, the PWAF has seen significant enhancements. A key addition is the Databricks AI Partner Dev Kit, offering over 15 AI-developed skills for tasks like integration patterns and telemetry instrumentation. This aims to allow coding agents to build against vetted standards, reducing manual implementation time.
New and expanded pattern guidance covers areas such as Clean Rooms, software-defined storage, and Marketplace apps. Existing guidance for fast-moving capabilities like Genie and Lakebase has also been refreshed.
Furthermore, the Firefly Analytics reference implementation, previously for Built-On partners, is now open-source. This provides working examples for authentication, security, scale, embedded apps, and AI, serving as a customizable starting point.
Partner Engineering in the AI Era
Databricks emphasizes that the evolution of partner collaboration is inherently tied to the AI era. The framework's AI-enabled guidance and tooling are designed to keep pace with both Databricks' product releases and the broader AI market.
The goal is to empower partners to build differentiated products faster, measure adoption impact, and unlock new growth opportunities. This includes enabling partners to leverage Databricks' platform more deeply to create unique solutions.
The innovation window is currently wide open, rewarding partners who build deep and differentiated offerings on the Databricks Lakehouse.
