The Partner Well-Architected Framework: What's New and What's Next

2026-06-17 · Source: Databricks · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure · Depth: Intermediate, short

Summary

The Databricks Partner Well-Architected Framework (PWAF), launched in February, provides AI-ready guidance for partners building on the Databricks platform. It unifies guidance across three core partner architectures: Built-On, Connected, and Data Collaboration, moving from static PDFs to dynamic, AI-enabled resources. Since its launch, PWAF has expanded significantly, introducing the Databricks AI Partner Dev Kit with over 15 AI-developed skills for coding agents, new pattern guidance for Clean Rooms, software-defined storage, and Marketplace apps, and refreshing existing guidance for Genie, Lakebase, and MCP server onboarding. Additionally, Firefly Analytics, a reference implementation for Built-On partners, is now open source, offering working examples for complex app development on Databricks. This framework aims to accelerate partner development and align with platform best practices.

Key takeaway

For AI Engineers or Architects building solutions on Databricks, the updated Partner Well-Architected Framework offers critical resources to accelerate development and ensure best practices. You should integrate the AI Partner Dev Kit into your workflows to automate routine integrations and leverage the open-source Firefly Analytics as a robust starting point. This approach helps you focus on product differentiation, ensuring your solutions align with Databricks standards and maximize GTM benefits.

Key insights

The Databricks PWAF provides AI-enabled guidance and tools to accelerate partner development across three core architectures.

Principles

• Architectures should be AI-ready by design.
• Standardize integration patterns for efficiency.
• Instrument solutions to measure adoption and DBU impact.

Method

Partners can use the PWAF architecture center, Databricks AI Partner Dev Kit skills, and open-source Firefly Analytics reference implementation to build, integrate, and validate solutions on Databricks.

In practice

• Point coding agents at the architecture center.
• Use Dev Kit skills for routine integrations.
• Clone Firefly Analytics for reference implementations.

Topics

• Databricks
• Partner Well-Architected Framework
• AI Partner Dev Kit
• Firefly Analytics
• Data Collaboration
• AI-enabled Guidance

Code references

• databrickslabs/firefly
• databricks-solutions/partner-ai-dev-kit

Best for: AI Engineer, AI Architect, Director of AI/ML

Related on AIssential

• Introducing the new Databricks Partner Program and Well-Architected Framework for ISVs and Data Providers — Databricks unifies partner programs and provides AI-ready architectural guidance to foster a robust data and AI ecosystem.
• Custom Agents now available on Databricks — Databricks Custom Agents streamline AI agent development and deployment with serverless infrastructure and integrated governance.
• From Data Analyst to Data Engineer: My 12-Month Self-Study Roadmap — Transitioning to data engineering offers deeper infrastructure understanding and career resilience against AI automation.
• Designing a Modern Enterprise Lakehouse with Databricks, DABs, and Medallion Architecture — Platform scalability in enterprise lakehouses relies on reusable engineering patterns and consistent deployment strategies, not just individual technologies.
• Accenture partners with Databricks on scaling enterprise AI solutions — Accenture and Databricks partnered to help enterprises scale AI by integrating data and deploying advanced AI agents.
• Fine-tune LLM with Databricks Unity Catalog and Amazon SageMaker AI — Integrate Databricks Unity Catalog with AWS ML services for governed LLM fine-tuning and end-to-end data lineage.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.