AI is rapidly becoming a creative partner for designers, reshaping how we work, think, and innovate. AI-powered design platforms are democratizing the field, enabling non-designers to produce professional-quality visuals and challenging traditional ideas of expertise.

With over 25 years of experience in design, I’ve witnessed the industry evolve dramatically—yet recent AI advancements have prompted deeper reflection on my role. Change can feel overwhelming, but it also opens new opportunities. Today, I explored two AI plugins that convert Figma designs into full-stack applications with just a few clicks. One is a familiar tool, the other a newcomer promising to push design toward hyperautomation and conversational interfaces. Let’s explore how AI is redefining the future of design.

First, the tools.

Figma by My Side

Figma needs no introduction for designers or developers. It’s become the go-to tool for creating, collaborating, and prototyping. With AI capabilities—either built-in or through plugins—Figma bridges the gap between design concepts and actual development.

What Does Builder.io Do?

Builder.io is a visual development platform that converts design into code, enabling teams to build fully functional websites and applications without deep coding knowledge. Its proprietary AI on Google Cloud imports designs and generates responsive code via Visual Copilot. Builder also integrates third-party LLMs like OpenAI and Anthropic Claude to edit content and optimize code.

How Can Lovable Help?

Lovable, launched in late 2024, is the new AI-powered tool we’ll focus on here. It acts as a full-stack AI software engineer and editing environment. Like Builder, Lovable can transform your design into a fully interactive application with authentication, using modern technologies like React, Tailwind, and Vite. It deploys on a custom backend powered by Supabase, an open-source database, making it a promising solution for designers and developers looking to streamline product delivery.

How Does it Work?

For the purpose of this article, we’ll assume you already have a design ready in Figma, ideally structured with auto layouts, since AI uses them to create a fully responsive application. If you don’t have a design file or lack the necessary design skills, you have several options: attend one of our training sessions, let Lovable generate a standard app template for you using Tailwind CSS, or use a plugin like Musho.ai to create a Figma design directly from a prompt. Lovable’s AI engine is pre-trained on thousands of app components and templates, allowing it to intelligently suggest architectures, libraries, and design patterns for a wide range of application types and styles.

Builder and Lovable now work together as a single integrated workflow, thanks to their recent partnership. This allows you to export your Figma designs seamlessly to either platform. While both aim to save you significant time by converting designs into fully functional code, each takes a unique yet complementary approach, giving teams flexibility while streamlining the transition from design to development.

The workflow itself is straightforward:

1. Select a frame in Figma.
2. Open the plugin and convert your design into code. The process takes a few minutes as the backend is automatically set up.
3. Choose your editor: Open it in Builder for a visual, drag-and-drop editing experience, or in Lovable, which allows further edits via prompts. Both tools are accessible without a login and offer free versions.

How Lovable Works in Figma

Once your design is rendered in Lovable, it features a multi-column layout with all UI elements in place. From this point, you’ll use prompts to interact with and refine your design. Simple adjustments, like changing colors, are handled flawlessly. However, for more complex or interactive code, it’s best to use contextual prompting strategies with smaller, incremental changes rather than large blocks of instructions all at once. Based on my (superficial) testing and reviewing, almost everything can be adjusted or fixed through this method. You can also integrate videos and images seamlessly. In the meantime, Lovable stores all your data in a Supabase backend.

The Lovable interface feels intuitive and engaging—almost like having a conversation with a developer. It provides real-time error detection and optimization suggestions, along with a library of tutorials, code snippets, and templates. That said, the execution can sometimes be inconsistent. Setup may take several minutes, with little feedback on what’s happening in the backend, and on one occasion, I was left staring at a blank (actually black) screen before the application finally loaded.

Both Lovable and Builder allow you to generate and refine code for free, but to deploy a fully functional application, a paid plan is required.

The UI of both engines look similar. Lovable (on top) works with prompts, whereas Builder.io (below) works with a visual WYSISYG editor and a code editor to allow for finegrained edits.

Conclusion

For designers and developers today, staying informed about AI isn’t just useful—it’s essential to staying relevant. Embracing these technologies, experimenting with them, and sharing constructive feedback are all key to shaping their evolution.

In this context, tools like Lovable (and alternatives such as Bolt or Sutro) are truly groundbreaking. They streamline workflows, letting us focus on design thinking, creative problem-solving, and technical architecture. By automating repetitive tasks, these tools free us to concentrate on the bigger picture: innovation and strategy.

I’ve shared my thoughts on the the future of (UX) Desig—feel free to check them out. I’d love to hear your perspective on these developments!