Research and development

Research on ontologies and AI agents

We work on four connected areas, from modelling knowledge to building agents that use it. Each project ends with something you can run: a prototype, a study or a product.

What we research

Four connected research areas

Ontologies and knowledge graphs

We model the concepts, rules and relations of a domain so that software and AI can reason over an organisation's data.

AI agents

We design and test autonomous and multi-agent systems that plan, use tools and carry out tasks.

Agents grounded in ontologies

We connect agents to knowledge graphs so that their answers can be traced to sources, checked and explained.

Applied AI for education

We study adaptive learning, skills modelling and assessment. This research feeds Subul, our learning platform.

How an R&D project runs

From the question to the handover, in three steps

1

Frame the question

We define the problem with you, review existing work and agree on what will count as a result.

2

Prototype and test

We build a prototype and test it on your data and your use cases.

3

Hand over

We deliver the code, the documentation and the results, and help your team take them over.