Where research meets reality.
Meet the AI Team
We develop the machine learning models that power IPRally’s search. Our work improves the models that power smarter, faster, more intuitive patent search.
🎯 What we’re here to do
At IPRally, AI isn’t just a feature, it’s the foundation.
We focus on enhancing the machine learning models that drive our patent search engine. That means tweaking existing models, discovering novel training strategies, and shipping improvements that go all the way to production.
Every team member wears both a research and engineering hat, taking ideas from concept to code to customer impact.
We focus on enhancing the machine learning models that drive our patent search engine. That means tweaking existing models, discovering novel training strategies, and shipping improvements that go all the way to production.
Every team member wears both a research and engineering hat, taking ideas from concept to code to customer impact.
🧪 How we work
Our approach is grounded in curiosity and practicality: We try things, evaluate quickly, and keep learning.
While our work is research-heavy, it’s always tethered to real-world results. We measure success in how well our AI helps users and we own the process end-to-end.
The vibe? We trust each other and like working on hard problems.
While our work is research-heavy, it’s always tethered to real-world results. We measure success in how well our AI helps users and we own the process end-to-end.
The vibe? We trust each other and like working on hard problems.
⚙️ Our tools & tech
We’ve built internal infrastructure that lets us focus on what really matters: improving ML.
We don’t just use tools, we shape them to amplify our work.
We don’t just use tools, we shape them to amplify our work.
IPRay
Our in-house tool to launch and resume training runs with a single command. No infra headaches.
Cloud GPUs
Automated, resilient setups (via IPRay) that just work.
Python, PyTorch, and friends
Flexible, familiar tooling that moves at our pace.
💡 What guides us
📚 What new teammates usually ask
Will I be doing real research or just tweaking someone else’s model?
You’ll be doing both. This isn’t a playground, but it’s not a factory either. You’ll apply existing methods, but are also free to chase original solutions that matter.
How does the team learn and grow together?
We reflect after every major experiment. It’s baked into our workflow. Everyone shares findings (even the ones that didn’t pan out) because our best breakthroughs often begin with a “failed” idea.
How close is the team to the product?
Very. Our work directly shapes the user experience. When the model gets smarter, our users see the benefit in no time. It’s incredibly motivating.
Where is the team based?
We’re mostly in Finland, but set up to work in a hybrid way.
What’s the working style like?
It’s a mix. Deep, focused work when you need it. Open, honest collaboration when it counts. The culture is remote-friendly and built on trust.
