Projects
A matter in one place. Colleagues write in the same chat, build on the same document, and share folders and instructions.
Egen, Swedish for our own, is a shared AI workspace for firms that hold other people's secrets. Documents, meetings and earlier decisions in one place. Built and run by a Swedish company, on servers in the EU that no US company and no US court can reach.
From raw material to finished work: you start with a contract, a recording or a question, and end with something your colleagues can build on.
A matter in one place. Colleagues write in the same chat, build on the same document, and share folders and instructions.
Upload a contract or a spreadsheet, ask, and get back a finished Word, Excel or PowerPoint file.
Record on Mac, Windows, iPhone or in the browser. Get a transcript with speakers, then minutes.
A saved instruction that runs on a schedule or when something happens, and leaves the result where you want it.
Give it a goal. Egen searches widely, reads the sources and writes a report with references.
Search the web without leaving the workspace. Every answer shows its sources.
Egen runs on Mac, Windows and iPhone, and in the browser.
You add people, assign seats and require two-step sign-in yourselves.
It reads only what you are already allowed to see, as you.
A law firm or an accounting firm cannot paste client material into a US chatbot and explain it later. Egen gives them the same kind of tool, on servers in the EU, with a log of every access.
Client material falls under legal professional privilege. Egen checks an agreement against your own template, lists the deviations and hands back a Word file your lawyers can edit.
Client books come with a duty of confidentiality. Egen reads a year-end workbook and returns the month-end report as a real Excel file, with the formulas still live.
The easy path routes your most valuable asset, your data and the models trained on it, through infrastructure and APIs you don't own. Quince builds the other path.
Before Egen, we built production machine learning for industry and the public sector, and we still do. We build the system and hand it over, put an engineer inside your team, or set up a private assistant on hardware you own. Either way, you keep the source and can run it without us.
We build the system and run it until your team can. Production builds on infrastructure you control, then ML Ops on a retainer designed to shrink.
Read more →A Quince engineer inside your team, by the hour, by the month, or for a fixed scope. Pipelines, model serving, monitoring, and the move off providers you cannot afford to depend on.
Read more →An open-weight assistant with your own documents behind it, set up on servers you own or on a single device, for when the material may not leave the building.
Read more →We're a young studio. As client projects close, they appear here as named case studies with verified numbers. Until then, our work notes show exactly how we think and build.
The AI HVAC system we build with our partner Mifimi: a physics model of comfort, learning per room, and the cheapest levers first. Now going into a faculty building in Belgrade.
A circular economy project with Wargön Innovation, funded by Vinnova: image recognition and near infrared spectroscopy fused into an automated sorting process.
The talent matching platform we built for ourselves and now run as a product, in production at Itmotive and Braha IT. What we built, and where we stopped the AI.
Quince AI is a machine-learning studio in Gothenburg, on Sweden's west coast, a region that quietly built automotive, manufacturing, and deep-tech engineering that had to work for decades. We bring the same discipline to AI: rigorous, dependable, and built to last.
Thirty minutes with an engineer. Tell us how many people and what is in the material, and we name the level and the price. If you need something built instead, the same call covers that.