Infinite canvas product search and decision making
Details
- External ID
- 45933456
- Source
- HN
- Company
- —
- Product
- Infinite canvas product search and decision making
- Website domain
- onton.com
- Launched
- Nov. 14, 2025
- Cohort
- —
- Upvotes
- 8
- Upvotes percentile
- 0.4421397379912664
- Tags
- —
- Fetched at
- Sept. 7, 2026, 9:25 p.m.
- Updated at
- Sept. 7, 2026, 9:25 p.m.
Description
Hi HN,We launched Canvas, an AI-powered visual workspace for furniture shopping: onton.com/surfaces. It's early, and we're trying to figure out if we're actually solving a real problem or just our own.Canvas lets you drag products from our 3M+ item catalog, upload inspiration images that auto-link to shoppable products, and generate AI room renderings. Basically trying to consolidate the nightmare of furniture shopping (Pinterest + screenshots + 15 browser tabs) into one place. The average person spends 79 days researching furniture across disconnected platforms. When I renovated my home, I had the same experience – inspiration everywhere, no way to see it all together or actually make purchase decisions.The interesting parts are probably the search and image generation systems feeding the canvas rather than the canvas itself. Making it infinitely scalable would be straightforward but the hard problems are upstream (product matching, AI rendering quality, search relevance across millions of items).Would love feedback if you try it, especially around whether this consolidation actually matters or if there are better ways to solve this problem.
Enrichment
- Theme
- 3D modeling and visual design tools
- Vertical
- Retail & commerce
- Function
- Search & retrieval
- Audience
- B2B
- AI stance
- AI feature
- Project type
- Commercial product
- Normalized one-liner
- product search and comparison tool
- Manually corrected
- False
Could you build this?
Partial The infinite canvas workspace UI is vibe-codeable, but scraping, indexing, vectorizing, and continuously maintaining a 3M+ furniture product catalog with visual similarity search is non-trivial.
What it would actually take: The stack pairs a canvas frontend (e.g., Fabric.js, PixiJS, or React Flow) with a backend indexing millions of SKUs using vector search (Qdrant/Milvus) powered by multimodal vision models (like CLIP or SigLIP). The core difficulty is the data engineering pipeline: scraping, deduplicating, standardizing metadata, and keeping pricing/inventory fresh across hundreds of retailers. It requires multi-modal ML engineering and distributed web scrapers running at scale.
Discussion
No comments on this launch.
Competitors
Other products that read as similar to this one — 72 launches clear the similarity bar, closest 8 shown.
Attention rank: #40 of 73 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 12 days after the earliest competitor.
- RetailInterior · ph · 2026-09-22 · 2 upvotes · similarity 0.44
- Slashspace AI · ph · 2026-06-11 · 340 upvotes · similarity 0.44
- Design Studio by Monday Merch · ph · 2026-09-11 · 92 upvotes · similarity 0.42
- BudgetPixel Design Agent · ph · 2026-09-26 · 1 upvotes · similarity 0.40
- Wonder · ph · 2026-04-30 · 247 upvotes · similarity 0.40
- Supanova · ph · 2026-09-29 · 1 upvotes · similarity 0.40
- Simple algorithm and color space to generate diverse skin tones · hn · 2026-08-04 · 627 upvotes · similarity 0.39
- Pixel Space Canvas · ph · 2026-09-29 · 1 upvotes · similarity 0.39
Other launches for this product
- No other launches for this product.