DesignKit 3.5: Supercharge Power Apps

Key insights

• DesignKit 3.5 brings AI-assisted building to Power Apps with an AI-powered Plan Designer that helps translate business needs into app logic and UI. 
  It reduces manual work and speeds up both citizen and professional development.
• Figma‑to‑Power Apps workflow supports exporting Figma screens to YAML and pasting them into Power Apps, plus importing screens back from Power Apps into Figma with dependency mapping. 
  The demo also previews multi‑screen export and improved font parity coming in v4.0.
• Reusable Components and asset libraries let teams copy, import, and reuse UI pieces, themes, and colors to keep apps consistent. 
  DesignKit also maps common control prefixes (like btn, gal, scr) to speed organization and maintenance.
• Data Model Management is integrated so makers can create, import, and edit tables inside the design flow and connect them to Dataverse. 
  This keeps data structure work close to UI and logic design.
• Vibe coding and natural‑language or sketch-based input let developers describe experiences and get generated interfaces and workflows, while Copilot offers real‑time help for formulas, content, and automation. 
  This bridges low-code with advanced customization.
• Responsive UI improvements and admin integration support enterprise needs with lifecycle, governance, and monitoring features. 
  The result is faster, more scalable app delivery with better control for IT and business teams.

Overview of the demo

The Microsoft-authored video demo, presented on the Power Platform monthly call on 16 July 2025, introduces DesignKit 3.5 and its role in Power Apps canvas development. The presentation, led by Lukas Pavelka, walks viewers through a practical workflow that connects design tools and the Power Apps environment. Consequently, the demo highlights how a Figma screen can be exported to YAML and then pasted into Power Apps, and vice versa, showing dependency mapping when importing back into Figma. This narrative frames DesignKit 3.5 as a bridge between designers and makers, enhancing handoff and iteration.

Moreover, the video includes a preview of upcoming enhancements in v4.0, which promises multi-screen export and improved font parity under the working name “Power Apps for Figma.” Therefore, viewers get both immediate techniques and a look at the roadmap. The demo emphasizes practical features like copy/paste snippets, control prefix mapping such as btn and gal, theme and color management, and component reuse. As a result, the presentation aims to show how teams can speed up app delivery while maintaining visual consistency.

Figma-to-Power Apps workflow and practical steps

In the demo, the Figma-to-Power Apps flow begins by exporting a single Figma screen into YAML, which captures layout and component definitions. Then, users paste that YAML into the Power Apps canvas where DesignKit recreates controls and mappings, simplifying what used to be manual recreation. The presenter also shows importing screens from Power Apps back into Figma with dependency mapping, which helps designers understand relationships between components and data. Thus, the process supports a two-way synchronization that reduces friction between design and implementation.

Additionally, the demo highlights small but important details like control prefixes and theme handling, which preserve naming conventions and visual rules across tools. For example, mapping prefixes such as scr or gal keeps galleries and screens identifiable after import, aiding maintainability. These conventions improve team collaboration because both designers and makers can reason about elements with the same vocabulary. Consequently, reuse of components becomes more straightforward and less error-prone across projects.

AI assistance and automation features

The video also covers AI-driven capabilities embedded in the plan designer and the broader development workflow, where natural language prompts guide design and logic generation. This approach, marketed as agent-assisted design, allows users to describe business problems in plain terms and receive suggested app structures, data models, and UI components in response. In turn, these suggestions reduce initial development time and lower the barrier for non-developers to contribute meaningfully. Importantly, the demo shows how integrated copilots can propose formulas, automation flows, and UI adjustments, speeding iteration.

However, while AI assistance accelerates workflows, the demo acknowledges tradeoffs around control and predictability, as automated outputs require human review and tuning. Therefore, teams must balance speed with governance to ensure the generated logic meets business rules and security standards. Furthermore, the presenter indicates that advanced users can still intervene with manual coding or refinement, which preserves flexibility for complex scenarios. Thus, the platform aims to blend automation with explicit oversight to support both citizen builders and professional developers.

Reuse, governance, and enterprise readiness

DesignKit 3.5 emphasizes reusable components, asset import, and consistent themes to drive enterprise-scale productivity and visual uniformity. By enabling teams to maintain libraries of components and data models, the tool reduces duplicate effort and enforces standards across applications. At the same time, the demo points to integration with governance tools so administrators can track lifecycle, security posture, and environment usage. Consequently, organizations can scale low-code initiatives while preserving compliance and oversight.

Nevertheless, the tradeoffs here involve balancing reuse with flexibility; overly rigid libraries can stifle innovation, while loose standards can lead to inconsistent user experiences. Therefore, the video suggests a middle path where curated component sets exist alongside options for extension and customization. Finally, administrators and design leads should plan governance policies that enable reuse but still allow exceptions when app needs demand them, keeping both speed and quality in focus.

Challenges, tradeoffs, and what to watch next

The demo makes clear that adopting a Figma–Power Apps workflow and agent-assisted design brings both benefits and challenges, especially around versioning, font parity, and multi-screen coordination. For instance, the preview of v4.0 signals progress on multi-screen export and improved font handling, but it also underscores the ongoing work required to achieve full parity between design and runtime. As a result, teams will need to manage expectations and plan for iterative improvements as the tool evolves.

Moreover, the presenter highlights the need for clear naming conventions, dependency mapping, and testing practices to avoid surprises when assets move between tools. In practice, this means investing some time upfront to establish naming rules and governance processes that the demo shows can be automated partly by DesignKit. Looking ahead, the combination of design fidelity, AI assistance, and governance creates a promising path, but it also requires careful change management to realize the full benefits without introducing risk.

Keywords

DesignKit 3.5, Power Apps design kit, Power Apps UI components, Power Platform design system, Power Apps development best practices, low-code UI kit, canvas app templates, Power Apps component framework