Treyspace turns an infinite canvas into a queryable knowledge graph. Draw, connect and organise information naturally, then ask questions across the whole canvas. Unlike ordinary RAG, Treyspace preserves the relationships, groups and spatial structure between elements, retrieves the relevant context, and highlights the sources behind its answers.
Hey Product Hunt — I’m Louis, the solo developer behind Treyspace.
I built Treyspace around a simple idea: when we organise information visually, the layout itself contains meaning. Boxes are grouped, arrows express relationships, and proximity often tells you what belongs together — but most AI retrieval reduces all of that back to isolated chunks of text.
Treyspace keeps that structure.
It’s an Excalidraw-based infinite canvas backed by a graph-aware retrieval system, so you can organise information visually and then query the canvas with AI. Relevant elements are retrieved using both their content and their relationships on the board, with the sources highlighted in the canvas.
I’ve also open-sourced the underlying Treyspace SDK under MIT for anyone interested in the retrieval side.
I’d particularly like feedback on where the graph-aware querying feels genuinely useful versus where ordinary semantic search would have been enough.