The four-day UCL AI Festival brought together students, researchers and entrepreneurs to explore cutting-edge research and real-world applications across healthcare, climate, quantum computing, robotics and more.
The programme combined a two-day hackathon organised with AI Engine with two days of research-focused presentations delivered in partnership with NVIDIA and HPE, and hosted at by British Land at One Triton Square, a joint venture with Royal London Asset Management. The event was a collaboration between UCL Innovation & Enterprise, who lead on strategic innovation partnerships including NVIDIA and HPE and the UCL London Office, who manage the relationship with British Land.
Hackathon
The festival opened on Saturday with a two - day hackathon that brought together more than 200 participants to prototype new tools and applications using the latest AI platforms. The open-ended hackathon, co-organised with AI Engine, gave teams just 24 hours to create and present a working demo that that would contribute to real world impact. Teams worked across healthcare, climate, robotics and public - service use cases, supporting rapid experimentation.
The winning team included three UCL participants - Lucas Lim, Joe Tan, Shashank Durgad all from the UCL Department of Computer Science, as well as Desmond Zee from University of Cambridge. For placing first, they received tickets and funded travel to NVIDIA GTC 2026 in San Jose this March
They built Sentrix, a next generation police force designed to protect organisations when deploying autonomous agent workflows. Using patrol swarms and investigation teams, the system is designed to intercept criminals, through observation, malicious intent detection and rapid escalation.
Joe Tan said: "Winning the hackathon felt surreal; it goes to show that the urgent problems attract the most attention when you have the conviction to go after them."
UCL students, Afthab Shiraz, Ashraya Poudel and Mehul Chourasia, also placed third in the competition, winning a Nvidia DGX Spark.
The hackathon was supported was supported by industry partners like Cooley, N47, Crane, Dawn Capital, powered by a large number of technology partners including NVIDIA, AWS and Anthropic, and leading European tech companies like Lovable, Prolific, Encord, Doubleword and Runware.
Research deep dives
The research programme opened with a welcome from Professor Geraint Rees, Vice-Provost (Research, Innovation & Global Engagement), setting the tone for two days focused on the technical foundations and societal impact of AI. Talks highlighted accelerated computing for research, sovereign AI infrastructure, healthcare digital twins, climate and polar observation, quantum computing and robotics.
Professor Geraint Rees said: "The UCL AI Festival shows both the power and responsibility of AI. From developers building new tools to researchers tackling unanswered questions, UCL is helping lead the field - advancing the science while ensuring AI delivers real-world impact for people and planet."
In her talk "From Lab to Market: The Synthesia Story," Professor Lourdes Agapito (UCL Computer Science) traced Synthesia’s journey from 3D vision research at UCL into a global generative-AI company with support from NVIDIA’s developer relations team.
Other talks likewise highlighted the real - world impact of AI, including:
- Professor Pontus Stenetorp (UCL Centre for Artificial Intelligence) described the efforts of his team at UK-LLM to train a large language model solely on British computational resources to enable AI reasoning in English and Welsh for public services.
- Dr Michel Tsamados (UCL Earth Sciences) highlighted how his team are developing advanced models to investigate the effects of climate change on the polar regions.
- Dr Chengxu Zhou (UCL Computer Science) presented recent work from the UCL Humanoid Robotics Lab on experiments developing AI systems to safely operate robots under changing and uncertain conditions.
- Professor Peter Coveney (UCL Chemistry) shared pioneering research that integrates quantum computing with GPU-accelerated supercomputing, using NVIDIA’s software framework CUDA-Q, to simulate the complex molecular mechanisms of life.
- Dr James Ruffle (UCL Queen Square Institute of Neurology) described ways that AI can assist healthcare professionals to interpret large amounts complex and diverse data to better diagnose and treat brain tumours.
- University College London, Gower Street, London, WC1E 6BT (0) 20 7679 2000
