Hive Edit: From a Manufacturer's API to Live Ads in One Conversation
Joona Nuutinen Hive Edit is a way to run a web store by talking to it. You describe what you want in plain language and it happens: a manufacturer’s whole catalog imported and categorized, every product enriched and on-brand, a launch campaign live on Google Ads, all from a sentence. It connects directly to the systems your products come from and exposes the entire store through a shared chat your team works in together.
The result is leverage that compounds, and it is built for small teams. When two to five people run everything between them, the mundane, time-consuming work piles up on all of them: imports, descriptions, data cleanup, the kind of thing that should have been automated long ago. Hive Edit takes that load. One person can do the work of a whole team, the routine runs itself, and the hours come back to people who are already stretched across every job. The name fits the idea: a hive is many workers building one shared structure, each doing its part, all contributing to the same store.
What is underneath it
Hive Edit runs on Hermes, an LLM scaffolding system in the same family as open agent frameworks like OpenClaw. The scaffolding is the part that lets a language model actually do things instead of just talking about them: connect to APIs, run tools, take multi-step actions, and check its own work along the way. That layer is becoming a commodity, which is good. It means the interesting work moves up to what you build on top of it.
Hive Edit is what we built on top for commerce. The connections to suppliers and to WooCommerce, the shared chat, the brand and catalog context, and the guardrails that keep a confident model from publishing something you did not approve. Hermes is the engine. Hive Edit is the store.
It starts at the source
A web store is only as good as the data behind it, and most of that data already exists somewhere upstream. The manufacturer has it. The distributor has it. The brand’s product portal has it. Hive Edit goes and gets it.
It connects directly to manufacturer and supplier APIs, reads the catalog at the source, maps it to your store’s structure, and brings products in: titles, attributes, variants, images, stock, pricing. A full catalog import is one sentence. “Import the new spring collection from this supplier, set the category, and apply our usual 2.3x markup.” It does the work, shows you what it is about to publish, and waits for the go-ahead. From there the same conversation can enrich every product, write the descriptions, and hand the finished range to your ad campaigns, without anyone opening a spreadsheet.
Everyone works at once
The part that makes the capability multiply is the shared chat. A small team does not have a copywriter, a category manager, and a marketing department. The same few people wear every hat. Hive Edit lives in a channel they all share, where anyone with access can talk to it and everyone can see what the others are doing.
One of you imports a new range while another enriches last month’s products and a third pushes a campaign live. The work that used to wait for someone to find a free afternoon just happens, in parallel, in plain language. The store becomes a space the whole team builds in together, and nobody is the single bottleneck it all has to pass through.
Because it is chat, the history is the audit trail. You can scroll back and see who changed what, why, and what the system did in response. One running record, no guessing.
What you can hand to it
The first things we built are the obvious ones: natural-language product creation and editing, mass imports, bulk edits across a category or the whole catalog, enriching existing products with detail, generating articles and buying guides, processing orders, and creating ads. But once the store is reachable through one intelligent interface, the list of things worth handing over keeps growing. A few of them:
Catalog and data
- Auto-categorize incoming products and build the attribute taxonomy to match
- Audit the catalog: surface products missing images, with thin descriptions, broken links, duplicate SKUs, or orphaned variants
- Keep stock and pricing in sync with the supplier, flag low stock, retire discontinued items
- Translate the whole catalog into new market languages in one pass
- Onboard a new supplier by mapping their feed to your fields automatically
Content and merchandising
- Generate SEO that actually ships: meta titles, descriptions, structured product data, image alt text
- Write comparison tables, sizing charts, and category landing copy
- Build bundles, kits, and cross-sell and upsell relationships from how products actually go together
- Curate collections and “new arrivals”, tune on-site search and synonyms
- Draft and answer product questions, turn recurring ones into per-product FAQs
Marketing and sales
- Create and validate ad campaigns from the live catalog (this is where Netura Campaign Manager takes over)
- Generate newsletter and launch content, abandoned-cart sequences, segment-targeted promotions
- Spin up seasonal sales, coupon codes, and flash-sale setups
- Produce social posts and captions for new products
- Publish the catalog out to other channels: Google Shopping, marketplaces, comparison sites
Operations and insight
- Process and screen orders, handle returns and refunds
- Answer questions about the business in plain language: “which products had the best margin last quarter, and which of those are we under-advertising?”
- Watch for anomalies in sales, stock, or feed health and raise them before you notice
- Keep up with compliance: VAT-correct pricing, energy labels, product safety disclosures for the markets you sell in
None of these are separate tools you have to learn. They are things you ask for.
What you get back
The interface is a sentence, so the learning curve is just knowing what you want. A new team member is productive on day one, because the store speaks their language.
That changes the math of what a small team can do. Enriching every product with real detail, translating the whole catalog into a new market, keeping a thousand descriptions on-brand: these become trivial, so they actually get done. The capability does not add up, it multiplies, because every contributor is working through the same intelligent layer at the same time.
And because Hive Edit reads the same catalog, brand guidelines, and supplier data that everything else runs on, the output stays consistent no matter how many hands are involved. The descriptions sound like your brand. The categories follow your taxonomy. The ads match the products. One source of truth, many hands, and most of your week handed back to you.
Where it fits
Hive Edit is built for WooCommerce, the platform we have run stores on for over two decades. It sits alongside the rest of what we make: Netura Campaign Manager for advertising, and our Google Merchant Center tools for keeping the product feed clean. Hive Edit is the layer that ties the store, the suppliers, and the team together into one conversation.
We are still building it, and we are shaping it around real stores. If you run a WooCommerce store, or you manage stores for clients and want to multiply what your team can ship, book a meeting and we will show you what running your store from a chat looks like with your actual products.