Socio-technical Systems

Distributed systems are a collection of independent computers connected via a network that function as a single, unified system to achieve a common goal. Examples of topics we work on:

• Using sensor data for the prediction of environmental risks or sporting performance
• Quantitative analysis of the "Verifier's Dilemma" in Proof-of-Work public blockchains
• Distributed financial environments and technologies

A framework where multiple autonomous AI agents interact, collaborate, or compete to solve complex tasks that are difficult for a single agent.

Federated Learning is a Privacy Enhancing Technology (PET) and a decentralised machine learning approach that trains AI models on edge devices (like smartphones), or local servers, without exchanging private data. Examples of topics we work on:

• Sharing of data to preserve privacy through Federated Learning
• Automated compliance verification of data sharing in decentralised settings (e.g. federated learning)
• Machine-readable policies for data sharing in decentralised (and centralised) settings

Data structures that represent a network of real-world entities (people, places, concepts, events) and define the relationships between them in a machine-readable format.

We investigate both Knowledge Graphs and Property Graphs as tools for responsible data management in AI (eg via Graph RAG).

We also work on ontologies and metadata to make data more Findable, Accessible, Interoperable, Reusable (FAIR) for AI.

We are dedicated to bridging the gap between technical innovation and ethics by building AI systems that are transparent, fair, and reliable. Our work spans topics such as algorithmic accountability, explainability of AI results, collaborative AI development, sustainability and regulatory compliance technology. Examples of topics we work on:

• Large Language Models and other AI methods to support defect detection in software engineering
• Augmenting LLMs via Graph RAG to mitigate hallucinations and improve trustworthiness
• Responsible data governance and management for AI
• Legal, Ethical, and Responsible uses of AI

We explore the intersection of machine intelligence and human cognition to create systems that amplify, rather than replace, human potential. Our work is rooted in collaborative intelligence, focusing on intuitive interfaces, explainable outputs, and adaptive learning environments. Through a multidisciplinary approach that combines computer science with behavioural insights, we design AI that respects human agency and addresses complex societal challenges through a user-first lens. Examples of topics we work on:

• Design and development of novel user interfaces, particularly through Augmented Reality or tangible user interfaces
• Empowering individuals by developing interfaces and technologies for greater data sovereignity online
• AI and data literacy initiatives