The posting
About Userlink
Userlink runs AI agents in production at several hundred Nordic webshops, including Bahne, XL-BYG, Coolshop and Bygma. The agents handle the customer dialogue before and after the purchase, and every conversation ends up as structured data on demand and purchase barriers. The team is based in Odense.
The brain
The agents are the arms. The next layer is a system that brings together every signal about a webshop and its customers: conversations on chat, email, phone and social, on-site behaviour, orders, returns and reviews.
The system identifies the moments where a sale is won or lost and intervenes while the session is still live. It scores answer coverage per product and closes the gaps, quantifies objections and unmet demand, and traces returns back to the expectation that broke.
Learning happens across all webshops, more than six million visits a month, where patterns become visible that no single shop has the traffic to see. Every intervention is measured against a control group. The patterns are shared, the data stays with each shop.
The hard problems are retrieval and evaluation of LLM output at scale, data isolation when patterns are shared across customers, latency in real-time flows, and attribution that holds up under scrutiny.
The stack
Java with Spring Boot, React and MongoDB, with language models and vector search on top. Moving the backend to C# and .NET is under consideration.
The role
You are one of the first engineers on the brain and own parts of it from idea to production: data model, retrieval, evaluation and the real-time flows that intervene in the session. You write the code yourself, ship it and follow it when it meets real customers.
The work is not split into tickets in advance. You help decide what gets built, and how, and your choices become the foundation the next engineers build on.
Decisions are made on arguments, not titles. Ideas get challenged regardless of where they come from, and the best one wins.
Why founding
You come in while the brain is taking shape. There is no finished architecture to follow, and the choices you make now are the ones everyone after you builds on.
That is why the role comes with equity. You help build the company, and you own part of what you build. In return we expect you to take responsibility like a co-owner, also when things are unclear or go wrong.
Profile
- Experience building and running SaaS systems in production
- Solid backend experience in Java or C#, and comfortable in React
- Has built LLM-based features and knows how to measure whether they work
- Owns what they build, also when it breaks in production
- AI-assisted development as part of the daily workflow
The person
- Dreams of owning and building a company
- Ideally has tried starting something and knows what it takes
- Thinks like a founder, not an employee
- Happily gives more than a regular job asks for, because it is fun
- Says things as they are, and can take the same in return
- Takes ownership and looks for solutions rather than excuses
- Curious, learns fast and has a positive mindset
What we offer
- Competitive salary
- Equity
- Short decision paths and a high degree of technical freedom
- Office at Storms Pakhus in Odense
Application process
We take the time it takes to find the right match.
- Introductory conversation
- Technical test
- Conversation about expectations and terms
Send a short description of something you built from scratch and put into production, and what you learned when it met real users, to [email protected].



