From cutting exam administration time by 90 percent to helping researchers bring medical software to regulators 60 percent faster, eight new AI for University Operations projects will use artificial intelligence to shape how the University of Cambridge’s administration works. The teams, selected from over 140 expressions of interest, will receive up to £10,000 each, alongside expert mentoring, to develop AI solutions addressing real operational challenges across the University.
The projects tackle diverse problems: automating 65,000 annual exam enrolment checks, analysing energy use to cut carbon emissions, streamlining research grant reviews across hundreds of funders, and making the University’s hidden Library collections more accessible. Each represents a practical test of how AI can support staff to make operations more effective.
“Through this programme, we’re giving professional services teams the opportunity to test how AI can meaningfully enhance the day-to-day running of the University,” said Jess Montgomery, Director of ai@cam. “The response to the call was fantastic with over 140 professional services staff expressing interest in the programme, with proposals that were clearly grounded in real operational challenges.”
The eight funded projects span departments across the University and address diverse operational challenges:
Funded projects
AI reply/chatbot for staff and student recruitment
Bennett School of Public Policy
This project will develop an AI-powered replybot and chatbot to handle high volumes of staff and student recruitment enquiries efficiently. The system will provide instant, consistent responses to frequently asked questions about job opportunities and postgraduate admissions, reducing manual email handling and improving response times. By automating routine queries, it will free up administrative capacity for higher-value work such as interview coordination and candidate support. Designed as a scalable model, the tool could be adapted across departments to support HR and student admissions more widely.
Team: Sarah Rosella, Rebecca Learn, Diane Coyle
AI-assisted Regulatory Compliance Technical File Generator for software as a medical device
Office of Translational Research, School of Clinical Medicine
By using AI to automate regulatory documentation, the Office for Translational Research and the School of Clinical Medicine are helping researchers bring medical software to patients faster. Its new “TechDoc co-pilot” will generate compliant technical files for Software as a Medical Device (SaMD), cutting preparation time by more than 60 percent and improving audit quality. The tool will be piloted in cancer risk modelling and neonatal care, with potential to support SaMD projects across the University.
Team: Reza Salek, Antonis Antoniou, Kathryn Beardsall, Raj Jena
Exploring AI in Research Grants Administration
Departments of Psychology and Engineering, UIS, Leverhulme Centre for the Future of Intelligence, MMLL, RSO, ROO, CRASSH, and Judge Business School
Research support teams handle a high volume of complex queries about funding opportunities, often searching across multiple funder websites and documents to find key eligibility details. This project will test AI tools such as Microsoft Copilot, ChatGPT, and Google NotebookLM to see how they can streamline these processes - from summarising funder guidance to checking eligibility and interpreting terms and conditions. The team will develop prompt libraries, usage guides, and practical recommendations to help research administrators work more efficiently and consistently, freeing up time for higher-value support.
Team: Petra Georgoulis-Hluzova, Peter Thomas-McEwen, Sabrina Yang, Hannah Tigg, Yvonne Martin-Portugues, Miyoung Kim, Sarah Lamont, Simeon Burke, Katerina Tsormpatzoglou, Russell Manning, Rowena Harvey
Automation of ‘exam standing’ process
Education Services
Each year, staff manually check around 65,000 exam enrolments against complex programme rules, a time-intensive process that can cause delays and errors. Using natural language processing (NLP), the project will convert these rules into a machine readable “rules library” to automatically validate student selections and flag issues for review. The system will cut manual checking by over 90 percent and speed up exam administration while improving accuracy and consistency.
Team: Jenny Blakesley, Jenny Green
AERO: Analytics for Energy-efficient and Resilient Operations
Department of Materials Science and Metallurgy
This project will develop an AI-driven energy analytics platform to help departments monitor, understand, and optimise their energy use. Using machine learning applied to gas and electricity data, the system will detect patterns, flag anomalies, and generate actionable recommendations for reducing waste and improving efficiency. A web-based dashboard will allow staff to visualise usage, benchmark performance, and model potential interventions. With high energy costs across the Department, the tool has strong potential to deliver substantial savings and carbon reductions, supporting the University’s sustainability goals and pathway to net zero.
Team: Jeanne Estabel, James Elliott, Xavier Montiel
Libraries Working Smarter with AI
Cambridge University Library
Cambridge University Libraries are launching a “Working Smarter” initiative to explore how AI can enhance efficiency and innovation across library services. The project will test AI tools for tasks such as metadata creation, text recognition, and search enhancement to improve access to “hidden collections” and digital content. It will also scope the potential for a library chatbot trained on internal documentation to assist users in navigating services and resources. Alongside technical trials, the initiative will engage staff across the library network in developing the skills and frameworks needed for the secure, ethical, and sustainable use of AI.
Team: Lesley Gray, Tuan Pham, Jay Pema
Smart-Review: AI-Powered Analysis of Grant Funding Terms
Cambridge Research Office
Managing thousands of research contracts each year, the Cambridge Research Office is developing an AI tool to analyse and compare grant funding terms across hundreds of funders. The system will deliver faster, more consistent reviews, highlight compliance risks, and support better strategic decisions. By automating routine checks, it will free staff to focus on complex negotiations and strengthen oversight of the University’s research funding portfolio.
Team: Claire Piffard, Sebastian Ashenden-Field
AI for Committee Management
School of Clinical Medicine and Governance and Compliance Division
This project explores how AI can streamline the administrative work of University committees, including minute-taking, drafting action lists, and summarising complex papers. Using secure and approved AI tools, the team will pilot AI support across a range of central and School committees. The project aims to reduce time spent on routine documentation while maintaining accuracy and confidentiality, freeing staff for higher-value work. Outputs will include practical guidance and recommendations to support wider adoption across the University.
Team: Jackie Hall, Michael Morgan, Claire Darracott, Barbara Bennett, Adam Russell, Jo Craigwood, Alistair Bochel, Emma Dollard, Caroline Evans, Emma Frampton, Angela Stratford
From ideas to implementation
Over the next six months, project teams will develop and test their prototypes with support from monthly AI Clinics. They will share their findings at a final showcase in May 2026, highlighting lessons learned and the potential for wider adoption across the University.
“This next generation of AI offers the potential to develop digital systems that adapt to us, that solve the problems we have. The best way to do that is to work closely with staff that deal with those issues on a day-to-day basis. AI for University Operations is about empowering colleagues to experiment responsibly and learn what works,” said Professor Neil Lawrence, Chair of ai@cam. “These projects show that when AI is applied to the real challenges our colleagues face, it can enhance efficiency and decision-making, allowing people to focus on the most meaningful parts of their roles.”
The AI for University Operations programme is part of ai@cam’s mission to drive a new wave of AI innovation that delivers real public value.
