illumi

illumi is a visual workspace designed to transform scattered group thinking into structured, deliverable outcomes. At its core, the platform provides an infinite canvas where multiple AI models coexist with real-time human collaboration. Users begin by dumping raw inputs—chat transcripts, sticky notes, meeting recordings, or fragmented documents—onto the canvas. The workspace then assists in organizing these inputs into clustered themes, relationships, and hierarchies. Instead of a linear chat interface, illumi employs a spatial layout where each piece of content can be moved, grouped, linked, or annotated. The AI models available on the platform serve distinct roles: one might summarize discussion points, another generates alternative frameworks, a third suggests visual hierarchies, and a fourth produces draft text or diagrams. Users can summon these models on demand, direct them to focus on specific regions of the canvas, and iteratively refine the output through conversation. The workflow is broken into five explicit phases: Input, Explore, Structure, Refine, and Output. During Input, users collect all source materials and place them onto the canvas as cards, text blocks, images, or links. The Explore phase encourages divergent thinking: AI models propose connections, related ideas, or counterpoints based on the existing content. Users can collapse, expand, or branch off subtopics. In Structure, the platform helps users build a shared mental model. They can create hierarchies, timelines, flowcharts, or argument maps. The AI can automatically detect structural patterns, suggest parent-child relationships, or identify gaps in logic. During Refine, users polish the content: AI models rewrite sections for clarity, adjust tone, expand on thin arguments, or consolidate red