ArchTech
A calmer way to search furniture across real catalogs, with review reasoning built in.
Article, Ikea, Rcwilley, Target, Worldmarket
301 products indexed
Beds, sofas, dining tables
Furniture assistant, not filter maze

Talk through the room. We’ll surface the shortlist.

ArchTech is built to feel more like a thoughtful shopping guide than a shopping form. Bring a rough idea, a strict budget, a style direction, or a retailer you trust. The conversation stays natural while the results stay grounded in real product and review data.

Current scope

Cross-store search with a renter-first point of view.

The site now compares real product data across multiple retailers and categories, then explains why a result fits instead of just handing back a pile of links.

301catalog items with shared product, style, size, and review structure.
5retailers flowing into one catalog the agent can actually search.
3categories live right now: beds, sofas, and dining tables.
Review-firstEvery recommendation surfaces a clearer reason, not just a score.

Conversation

Start with whatever feels easiest: a price ceiling, a product type, a platform, or just the mood you want the room to have.

You can be direct, vague, or change your mind halfway through. The assistant should move with you.

Shortlist

Matches appear here once the conversation has enough signal. Cards lean on actual catalog data, not invented summaries.

ArchTech is still an evolving prototype, but the core loop is already live: conversation, shortlist, reasoning.
Back to conversation