About this challenge
Build a working prototype of an AI powered collaborative workspace where teams can meet, talk, and think together on a shared visual canvas. As the conversation happens, AI understands the discussion and turns ideas, decisions, questions, and relationships into visuals in real time, bringing the meeting and the canvas together in one seamless experience.
This video explains what problems currently exist in an existing app, solve them and you win.
What to build
- Build a collaborative meeting workspace where teams can have audio/video conversations directly inside the canvas.
- Think Google Meet + Miro + AI - the meeting and the visual workspace are connected instead of being separate tools.
- While people talk, the AI should understand the conversation and automatically build the canvas with ideas, topics, decisions, questions, tasks, diagrams, and other useful visuals.
- Participants should be able to talk, see each other, and interact with the canvas simultaneously.
- The canvas should update in real time as the discussion evolves, keeping the entire conversation and visual context connected.
- Users should be able to tell the AI what to change naturally, such as “group these ideas,” “move this under the problem,” or “visualize this.”
- Support collaboration features such as shared cursors, seeing where other participants are working, and Follow Me mode so everyone can follow the presenter.
- The workspace should be persistent, so the team can leave a room and return later with the conversation context and canvas still intact.
- The goal is to make the meeting itself produce a living visual representation of the team's thinking, without someone having to manually build the board.
Who should enter
- Open to developers, AI engineers, full-stack builders, product designers, indie hackers, and small teams.
- Ideal for builders interested in AI, real-time collaboration, video/audio communication, canvas interfaces, and agentic products.
- Any tech stack or AI model is allowed, choose whatever you think is best.
- You don't need to build the entire vision. Focus on proving the core experience: people talk → AI understands → the shared canvas builds itself.
How it works
- 1
Enter the challenge
Register to unlock the brief and get started.
- 2
Build your solution
Ship a working solution against the challenge brief.
- 3
Get reviewed & rewarded
Top submissions are evaluated against the scoring rubric.
Challenge task
Build an AI Collaborative Visual Workspace
Advanced~40hBuild a working prototype of an AI powered collaborative workspace where teams can meet, communicate, and think together on a shared visual canvas. The experience should bring video communication, real time collaboration, and AI powered visualization into one seamless workspace.
What to build
Expectations
Skills
Resources
Enter the challenge to unlock resources
- Deadline
- 20 days after applying
- Attempts allowed
- 10
How we evaluate
- Product Experience: How intuitive, polished, and seamless the overall meeting and canvas experience feels.
- AI Intelligence: How well the AI understands the conversation and turns it into meaningful visuals and structure.
- Real Time Collaboration: How smoothly multiple people can communicate, interact, and work together on the shared canvas.
- Visual Quality: How effectively the product transforms ideas and discussions into clear, useful, and engaging visuals.
- Adaptability: How well users can guide the AI and reshape the canvas using natural language.
- Technical Execution: Code quality, architecture, performance, reliability, and overall execution.
- Innovation: How creatively the submission pushes the idea beyond a traditional video meeting or whiteboard.