Walk into any retail back office. Open a supplier's product management workflow. Somewhere in there, someone is typing a product description from scratch. Copying specifications out of a PDF datasheet. Assigning categories manually, one SKU at a time, because that is simply how it has always been done.
This is not a startup problem. It is not a legacy technology problem. It happens at every scale of retail, in single-location stores, in national chains, in supplier organisations managing tens of thousands of SKUs. The tools have evolved over the decades. The underlying work has not.
Product data enrichment, the creation, standardisation, and ongoing maintenance of product information, is one of the most time-consuming operational tasks in retail. It is also one of the least visible. It never appears in a pitch deck. Nobody celebrates when it goes right. But when it falls behind, the consequences are immediate and tangible: incomplete listings, mismatched specifications, products that never make it online because there was not enough time to write them up properly. Revenue that simply does not happen.
This is the problem we are starting with.
Vesko is building Applied AI into the Retail OS
Beginning with product data enrichment.
The goal is not AI for its own sake. Vesko was building toward an AI-first Retail OS before applying AI to retail workflows became the obvious thing to say. What we are doing now is making that concrete, starting with the workflow that affects every retailer and every supplier on the platform, every single day.
The Applied AI layer will generate accurate product descriptions, standardise specifications across suppliers, and assign categories at a speed and scale that manual processes cannot come close to matching. All of it runs directly inside the Retail OS connected to the inventory retailers and suppliers are already managing in Vesko.
No separate integration. No new system to learn. No technical capability required from the retailer. The intelligence lives inside the platform they already use.
Why we started here
We chose product data enrichment deliberately. It is not the most glamorous problem in retail technology, and that is precisely why it has not attracted serious automation until now. The tools that exist today mostly help people organise manual work more efficiently. We are building something that removes the manual work from the process entirely.
The vision behind the Retail OS has always been one platform, everything a retailer needs to manage, operate, and sell, both offline and online, without stitching tools together or running parallel systems. Applied AI is not a new direction for Vesko. It is the completion of that vision: intelligence embedded into the operational layer, working in the background, so the retailer does not have to.
Product data enrichment is the first manual workflow we are replacing. It will not be the last.
What comes next
Our first Applied AI MVP is targeted for end of 2026. This is the beginning of a longer roadmap. Retail operations are full of repetitive, time-consuming tasks that have been accepted as a permanent cost of running a physical or omnichannel business. We do not accept that framing.
More detail on the roadmap, and the specific workflows we are tackling next, will follow as development progresses.
We are hiring AI engineers
Building this requires engineers who want to work on problems that are real, operational, and affect how retailers and suppliers run their businesses every day.
Not demos. Not wrappers. Real operational infrastructure, applied to a sector that has been underserved by serious technology investment for too long.
If that is the kind of work you want to do, we want to hear from you.
Open applications: hello@vesko.fi
.webp){kind=small}
.webp){kind=small}
.webp){kind=small}