Education should feel intuitive for teachers, for students, and for schools. We believe technology can remove the friction that slows learning down, giving every learner the space to grow and every educator the support they deserve.
Designing tools that support teachers, empower students, and make learning more equitable.
A world-class product built by a world-class team
Gradr is powered by handpicked talent from Sweden's top universities and companies. We move fast, build with purpose, and care deeply about the future of education.
Great teaching shouldn't be held back by outdated systems.
Across the world, teachers spend countless hours grading, organizing materials, and managing workflows that could be automated. Students, meanwhile, often study without guidance or clarity on how they're improving. We saw an opportunity to fix both sides of the learning experience — by building a platform that connects them.
Gradr brings teaching and studying into one unified flow, creating a shared space where progress becomes visible, feedback becomes actionable, and time is used where it matters most.
How we work
Every feature in Gradr comes from something a teacher asked for. We sit in classrooms, call after releases and rebuild what we got wrong. That is why the platform feels like a teacher’s tool rather than an AI demo.
Gradr sets no grades and passes nothing on to a student automatically. We prepare the material — the teacher reads it, edits it and decides what is returned. That order is not a setting, it is how the product is built.
Everything is stored with Hetzner in Finland, AI processing stays inside the EU, and student texts are never used to train models. We process student data as a processor under a data processing agreement (GDPR art. 28), on the school’s instructions.
Our average support response time is 90 seconds. We come out to the school, train on site and fix what breaks the same day when we can. A small team answering for itself is the whole point.
We go all the way for the people who use Gradr. Always, no matter what.
If something breaks in the middle of a marking week, we put down whatever we are holding. If a school needs help on site, we get on the train. If a feature is missing, we build it — and if you haven’t heard back from us yet, it is because we are already on it.
We don’t leave a teacher hanging, and we never hide behind something being complicated. This is not a support tier we sell — it is how we think about the work.
"It’s wonderful to see the teachers’ positive engagement. They have appreciated both Gradr’s affirming feedback on the qualities students demonstrate and its forward-looking suggestions for how students can continue to develop their knowledge and skills – it’s motivating for the students."
We run a TypeScript monorepo that powers an AI-driven education platform. We're obsessed with type safety, static analysis, and using LLMs to ship faster. If you don't believe AI can make you significantly more effective — this isn't the right place.
Table stakes:
- Comfortable in TypeScript. Generics, discriminated unions, type narrowing — not just aware of them.
- Real experience with PostgreSQL. You understand normalization and know that a well-designed schema is the foundation everything else builds on.
- You use LLMs as part of your actual workflow. Copilot, Cursor, agentic coding — whatever gets the job done.
- You care about correctness. We lean hard on strict TS, linting, static analysis, and CI so we can move fast without breaking things.
On top of that, you lean into at least one of these:
- LLM expert — You deeply understand how to build with language models. Prompt engineering, structured output, tool calling, evals. You push what's possible with AI in our product.
- Product-minded developer — You have strong design sense and product instincts. You can go from problem to shipped feature making good UX and scoping decisions along the way, not just executing specs.
- Leverage engine — You make the whole team faster. Better abstractions, better tooling, better CI pipelines, better linting rules. You're obsessed with automating correctness and multiplying developer output.
Day to day:
- Ship features across the full stack — SvelteKit frontend, Node backend, PostgreSQL.
- Work with LLMs daily. This is an AI company, not a company that bolted on an AI feature.
- Own problems end-to-end — from schema design to deployed feature.
- Write code that compiles clean. We'd rather catch bugs at build time than in production.
Stack:
Try Gradr with your own class!
Book a short demo – we'll show you how it works, you decide if it fits.
Book a demo