Applied AI Systems
How to get models out of the lab and keep them working: deployment, monitoring and evaluation in live settings.
- MLOps
- Evaluation
- Reliability
Research
Research that has to be used
We work with universities, hospitals, companies and public bodies on applied AI research. We bring engineering and data expertise, and we choose projects where the results are meant to be deployed, not just published.
- Built to be used
- We design for deployment from the start, not as a follow-on.
- Rigour
- A clearly framed question and an evaluation plan before any code.
- Responsibility
- Data governance, privacy and local ownership are part of the design.
What partners can expect
How we work on collaborative research.
Research focus areas
Four areas where we have both the expertise and an interest in collaborating.
Generative AI (RAG & Agents)
Retrieval and agent-based systems that can be evaluated for accuracy and kept within safe limits.
- RAG
- Agents
- Guardrails
Responsible AI & Governance
Fairness, transparency, privacy and regulatory compliance treated as design requirements from the start.
- Governance
- Privacy
- Fairness
AI for Public Good
AI applied in health, education and public services, where the results have to work for the people affected.
- Social good
- Sustainability
- Public sector
Collaboration principles
We look for partnerships that produce methods, systems and evidence someone will actually use. Publication matters, but it is not the point. Where it fits, we co-supervise students, feed research into our teaching and share infrastructure.
What we look for
- A clearly framed problem with a real user or beneficiary
- A realistic route to deployment and evaluation
- Attention to ethics, privacy and governance from the outset
What we bring
- Research experience combined with engineering depth
- MLOps, systems design and the discipline to deliver on time
- Training and knowledge transfer so partners can maintain what we build
Ways to work together
Academic Partnerships
Joint projects, co-supervision of students, shared datasets and co-authored publications.
Industry Collaboration
Developing and testing AI methods with companies that intend to use them.
Grant & Consortium Work
Shaping proposals, leading technical work packages, and handling evaluation and delivery.
Start a collaboration
A short note is enough to begin. We will reply by email.
Collaboration request
Tell us about your interests, the context and any timeline. If you are responding to a funding call or have a brief, mention it.
What to include
- A paragraph on the problem and who it affects
- Your timeline and any constraints
- Datasets, evaluation criteria or stakeholders already involved
What we are looking for
Partnerships that end with a system in use, a method that has been tested against real data, and partners who can carry the work on without us.
Relevant grants & funding
Open funding calls from UKRI, Innovate UK and the EU Funding & Tenders portal, ranked by how relevant they look to AI, data and digital skills. If one fits your work and ours, get in touch.
Browse all opportunities: UKRI Funding Finder|Innovate UK competitions|EU Funding & Tenders portal|GOV.UK Find a grant