Agentic AI
Not another chatbot: software that sees a task through, picks up the tools it needs along the way, and stops where a human decision belongs. It matters to us because it is the point where AI touches a process instead of commenting on it.
Not another chatbot: software that sees a task through, picks up the tools it needs along the way, and stops where a human decision belongs. It matters to us because it is the point where AI touches a process instead of commenting on it.
Running the computation where the data is born, instead of shipping it to a distant data centre. On a factory floor that means deciding in milliseconds even when the internet link drops. Our IoT projects already work this way, and it is worth pushing further.
We do not work on it, we watch it. The day a quantum computer breaks today's cryptography, the systems we are building now will already need algorithms that hold. That is why we follow it today, not for the science fiction.
Training a model without moving the data from where it lives, using federated learning and differential privacy. It matters wherever data cannot leave the building: healthcare, banking, production data. Exactly the sectors we work in.
Emerging technology is not a thought experiment for us. It is the raw material for products of our own: ideas we validate on the market and turn into companies.
The method, in short: we start from a real problem, build the minimum needed to test it, and scale only what works. It is a process we sharpen with every project.
Let's talk about it. Message us on WhatsApp and you will hear back within minutes.
An open space where research, teaching and prototypes live side by side: a physical place to experiment with students, companies and the wider community, and turn ideas into things that work.
