💡 Inspiration
While using Adobe Express, we found the tool incredibly powerful but challenging to navigate as non-experts. We wanted to make Adobe Express accessible and intuitive, enabling anyone to create beautiful designs—even without prior design or technical expertise.
⚠️ Challenges
One of our major technical hurdles was integrating Gemini AI with Adobe Express through the Model-Context Protocol (MCP). Ensuring that Gemini accurately understood and correctly executed context-aware function calls required significant experimentation, precise context capturing, and meticulous prompt engineering. Overcoming this allowed us to deliver a smooth, responsive, and intuitive user experience.
🎨 What it does
Buildr is an intelligent AI assistant directly integrated into Adobe Express. It allows anyone to easily create professional-grade graphical designs with plain words—translating creative imagination into reality without requiring any in-depth technical expertise.
⚡ How we built it and how it works
As you give your creative commands, Buildr translates your words into graphical instructions, explaining its mental process step by step and carrying out design commands directly through Adobe Express.
Buildr functions on an intelligent workflow that is designed to ensure precise and effective completion in Adobe Express:
Context Building: Upon receiving a user input, Buildr executes a Breadth-First Search (BFS) algorithm to fully capture the current state of the Adobe Express document—dimensions and all objects on the page and their positions.
Contextual Prompting: This contextualized picture is mixed with your prompt to provide Gemini rich context by a model-context protocol that ensures accurate and proper function calls.
AI-powered Execution: Gemini executes this contextual prompt, reflectively streaming its reasoning process step by step to provide full transparency.
Context Building: When the user enters a command, Buildr employs a Breadth-First Search (BFS) algorithm to get a complete snapshot of the state of the Adobe Express document, including dimensions and all objects that are currently on the page and their locations.
Contextual Prompting: Your prompt is paired with this detailed snapshot to offer rich context for Gemini, utilizing a model-context protocol that allows for precise and relevant function calls.
AI-directed Execution: Gemini converts this situational request, formalizing its cognitive process stepwise for absolute openness.

Straight Access to Adobe Express: Gemini formulates precise calls to function exactly as they occur, with straight execution through Adobe Express Document APIs (code.js environment), without delay to updating on-screen graphical items.
Technical Features:
Breadth-First Search (BFS) guarantees that Gemini has the latest snapshot of the document always. We ran a bfs to search through all the nodes (objects) on the canvas and uploaded it to gemini as context so it has a rough idea of what the users canvas looks like before it decides to call any functions.
Model-Context Protocol provides Gemini AI full context so there can be precise and meaningful calling of functions.

Instant visual outcomes with Adobe Express Document APIs create a smooth, high-efficiency creative workflow.
💼 Business Impact & Broader Adobe Integration:
Buildr makes Adobe's creative force accessible to users of all skill levels, dramatically expanding market reach and user engagement.
Greater Potential for Adobe Integration:

Summary
Buildr isn't designed to replace traditional methods but rather to complement and enhance them. By enabling natural language interaction, Adobe can empower a broader audience—particularly those without professional design experience—to easily achieve impressive results.
In short: Buildr makes creativity accessible. Just describe it, and Adobe Express brings it to life.


Log in or sign up for Devpost to join the conversation.