
ORO is building an AI shopping assistant, but the part you can actually use today is a developer platform for testing and submitting shopping agents against a controlled benchmark of real product data.
ORO is best understood as two products at once, but only one of them is publicly usable today. The public website positions ORO as a shopping assistant for ordinary users, while the docs and GitHub repo show that the live product is a developer system for testing shopping agents on a controlled benchmark. The real target user today is a developer or researcher building shopping agents, not a normal shopper. The consumer-facing promise exists, but the only public consumer action found was joining a waitlist. The developer surface is live: docs, quickstart, public monitoring API, active suite endpoint, leaderboard, local test harness, and GitHub repo. The ordinary-user shopping assistant remains incomplete or not publicly available.
The real target user today is a developer or researcher building shopping agents, not a normal shopper. The consumer-facing promise exists, but the only public consumer action found was joining a waitlist.
No reliable public source found for user count, customer count, revenue, or profit. The best public proxies are weak technical signals such as repo activity and rapid release iteration.
Closest alternatives include Amazon Rufus, Perplexity Instant Buy, Shopify Shop’s AI shopping assistant, Constructor’s AI Shopping Agent, and Instacart’s Cart Assistant. ORO appears stronger on openness and benchmark reproducibility, while weaker on public consumer readiness and market proof.
ORO is a real project with live technical infrastructure, but it is not yet a ready-to-use shopping assistant for ordinary users. It matters most if you are a developer building shopping agents in a controlled benchmark. Its main weaknesses are product readiness, public commercial proof, and transparency about team/company structure.