Our story
We set out to build AI for ecommerce. We ended up building its core.
Cernel was founded in Aarhus in 2023. We started out building AI that shops and recommends, watched it break on messy product data, and turned to the foundation underneath every channel.
- Founded
- 2023
- Based in
- Aarhus, Denmark
- People
- 40+
- Seed round
- €4M
Why it matters to us
We have worked inside ecommerce teams.
We have seen talented people spend their days cleaning spreadsheets instead of building campaigns. We have watched product launches slip because a file was missing attributes, and seen revenue sit in warehouses while teams rewrote descriptions by hand.
That never felt right.
Working with large language models, the founders went through a lot of trial and error to find the right way to work with AI. What made the output noticeably better every time was the same thing: deep, specific context from well-structured product data.
So we tried the ambitious things first: AI that shops for you, recommends products and runs your store. At a small scale it worked. At real scale it broke, because underneath every AI feature is product data, and in ecommerce that data is messy, inconsistent and structurally misunderstood.
So we pivoted. Instead of building AI on top of broken foundations, we started building the foundation itself.
The name
Cernel means core.
In software, the kernel is the core part, the part everything else rests on. For ecommerce, that part is product data, and it is what we build.
We swapped the K for a C, for cognitive.
Why now
Shoppers stopped typing keywords. Agents started doing the shopping.
People now search the way they talk, with a whole intent in one sentence, and AI agents increasingly find, filter and pick products for them. Both only work when the product data is structured and machine-readable.
We estimate a product now needs more than 100 times the structured attributes it used to, soft ones like good for hiking included. Teams are already trying, but each department works alone in spreadsheets and feed tools, so the work is duplicated and lost. That much structured data can only come from a system built for it.
What the product data has to answer
- PatternNot needed
- ColourNeeded
- Product typeNeeded
- Good forNot needed
- SeasonNot needed
Two keywords. Two fields are enough.
Illustrative example.
What we are building
The product data foundation under every channel.
Cernel sits between raw supplier data and every sales channel, structuring and enriching product data at scale. It shows its reasoning, keeps its sources and improves with feedback, so people and AI can work on the catalogue together.
Catalog › Women › Knitwear and Tops
New this week 5 products
ProductMaterialFitCareDescriptionStatus
Ellen Rollneck, StripeImported
Vigge Waffle, MossImported
Sol Shirt, WhiteImported
Tove Wrap Top, BrownImported
Mira Blouse, RoseImported
- 01
Find every gap.
Cernel shows what each category and market is missing, and which of those gaps cost you searches.
- 02
Fill it by your rules.
Agents fill every field from your data, your lists and the maker's own pages, in every language you sell in.
- 03
Send it where you sell.
You choose which values go live automatically and which get a review step. Then the product goes back to your shop or PIM, into a Google Shopping feed, or out as a file or over the API.
Milestones
From Aarhus, one step at a time.
The steps that shaped Cernel so far, from the first team in Aarhus to the one building it today.
All news2023
Founded in Aarhus
Cernel starts in Aarhus with a team of university students, award-winning mathematicians and NLP experts.
The pivot
From AI on top of messy data to the foundation
Underneath every AI feature is product data, and in ecommerce it is messy. So instead of building AI on top of it, Cernel starts building the product data foundation itself.
February 2026
€4M in seed funding
Seed Capital leads a €4M seed round, joined by new angels and all existing investors.
Read the announcementToday
40+ people, and counting
More than 50 million values enriched for customers so far, and 4,234,324 more every month, across markets.
The people
Built by people who have done the work, and people who build the best software.
- Values enriched for customers so far
- 50M+
- Values enriched in a typical month
- 4,234,324
The monthly figure counts up at its average pace while you watch: an illustration, not a live feed.
Engineers, mathematicians and competitive programmers, working next to SEO experts, ecommerce specialists and master data people who have onboarded products by hand themselves.
People at Cernel have worked at
Backed by Seed Capital and a group of serial entrepreneurs and angel investors.
Join the teamWhat we care about
Behind the scenes, too much of ecommerce runs on repetitive data work.
- We care becausePeople's time matters.
- We care becauseGrowth should not be blocked by formatting issues.
- We care becauseTechnology should remove friction, not create it.
It is how we work together, too. We build AI, and with AI. Everything starts and ends with our users, and we keep the ego low and the bar high.
How we workRead on
The next chapter is still being written.
Careers
Open roles in Aarhus, how we hire and how we work.
News
Funding, events and what we are learning about product data.
- Sponsored · ExhibitedStora Ehandelsdagen 2026
Cernel serves the coffee at Stora Ehandelsdagen 2026 in Stockholm
- Exhibited · SpokeE-handelskonferencen 2026
Cernel exhibits and speaks at E-handelskonferencen 2026
- Sponsored · ExhibitedDigital Clubhouse Summit 2026
Cernel sponsors the Digital Clubhouse Summit 2026 in Manchester