By Tomomi, Overseas Team, orosy
The short version. orosy just launched a headless wholesale marketplace. I run a Shopify B2B store that exports Japanese products overseas, so I did the obvious thing: I connected the new API to my own store and set it up — not by writing code, but by talking to my AI the way I already do. These are my honest first notes.
Key takeaways
- orosy released a headless marketplace — meaning it built the engine (an API), and you connect it to the store, app, or system you already run and build on top of it however you like.
- You don't have to be technical. I set it up by talking to my AI (Claude, ChatGPT, Gemini) in plain language — "list these kinds of products on my store" — the way I already work.
- I used it to start listing several hundred overseas-loved Japanese stationery products on my Shopify store, and set prices and pack sizes freely along the way.
- When some product photos had Japanese text baked in, I simply told the AI to exclude images with Japanese text — and it did, instantly matching my store's taste.
- From creating the API account to products appearing on my store took under an hour. The most interesting part: you can list before you stock, and only buy inventory once something actually sells.
First, what "headless" actually means here
A normal marketplace is a website you sell on. A headless one is different: orosy built the engine — an API — and left the storefront to you. You plug it into the store, app, or inventory system you already run, and from there you can even build the extra pieces only you need. The marketplace you can see is just one storefront on top of that engine; yours can be the next one.
"API" sounds hard. It wasn't.
When people hear "API," they picture something that needs a developer and a high tolerance for technical setup. That was my worry too. It turned out to be the opposite, because the whole thing can happen inside the AI you already use.
Here's the actual flow:
- Sign up and issue an API key — Create an account and generate an API key — about as involved as signing up for any app or social account. A few clicks, no coding.
- Tell your AI what you want, in plain language — Hand the key to the AI you already talk to — Claude, ChatGPT, Gemini — and just say what you're trying to do: "Use this API to list products on my store," "I want products like these," "show these on my shop." No syntax to learn.
- Watch it set up your store — It connects, pulls the products you described, and sets up the listings, with images, prices, and details in place.
What I actually did
I used the API to start listing several hundred Japanese stationery products — the kinds of brands that already have a following outside Japan — on my Shopify store. The setup really was a conversation: I described what I wanted, and it happened.
Along the way I could set prices and pack/lot sizes freely. That mattered more than it sounds — Japanese products often sell in fixed units, and being able to say "sell this in sets of N" in plain language kept my listings clean.
Then a real, small problem came up — and this was my favorite moment. My store is for overseas buyers, but some of the product photos had Japanese text baked into the image. I just told the AI: exclude the images that have Japanese text. It did, immediately. In one sentence, my listings looked native to my own audience. That's the part that made me stop and think — I was shaping the store to my taste by describing it, not by clicking through hundreds of images.
"But the catalog is huge — won't it be a mess?"
The catalog already runs deep across Japanese products, and I think the natural fear is: searching that is tedious, and if I import everything I'll drown in noise. Here's the reassuring part — you don't sift through it. You tell it the store you want. You can even point it at a store you admire, or your own existing shop, and have it match that direction. You curate by intent, not by scrolling.
And it's fast. From making the API account to products live on my store was under an hour. Reviewing afterward and fixing little things takes some care — that's the real work — but the setup itself is nearly instant.
The benefit I keep coming back to
Traditional retail makes you buy inventory before you can sell it. With this, I can list first, watch how my customers respond, and stock only after something actually sells. That quietly removes the biggest cost of carrying a shop — money tied up in inventory that might not move.
And because it's an API — an engine — you're not limited to what I did. You can build your own features on top of it; the uses are open-ended. The catalog and reach keep growing, too, in the direction of sourcing widely and simply.
I'm going to keep posting these field notes as I actually use it — the honest version, discoveries and all. Worth checking back.
FAQ
Do I need to be technical to use it?
Honestly, no. If you can sign up for a social media account, you can do this — creating an account and generating an API key is just a few clicks. From there I set everything up by talking to my AI (Claude, ChatGPT, Gemini) in plain language, describing the products and the store I wanted, not writing code.
The catalog is enormous — how do I avoid importing junk?
You don't browse the whole thing. You describe the store you want — even point the AI at a shop you're modeling yours on — and it reflects that intent. Curation happens by describing what fits your shelf, not by scrolling through everything.
Can I really list products before buying inventory?
Yes — that's the part I find most useful. You can list first, see what your customers respond to, and stock only once something sells, rather than paying to hold inventory up front.