DubLINKAI in Healthcare LLM-athon
Team 10 won the first DubLINK.
Six teams presented, judged on the failures they found, teamwork, creativity, the questions they asked and clarity.
- First place
Team 10
- Narod Daldal
- Sheila Zaragoza
- Jean O'Sullivan
- Apoorv Purohit
- Gabor Zilahi
- Rabiat Muhammed
- Second place
Team 9
- David O'Regan
- Pat Conroy
- Charles Wesley
- Sergey Avdeychik
- Conor Kenny
- Third place
Team 2
- JP Fitzpatrick
- Marina V.
- Cesar Alba Moreno
- Anna Nikitchenko
The day, in pictures.
Photo credit: Tamás Tiszai-Szűcs, Ezi Ozoani, Ahmad Albarqawi and Sebastián Cajas.
A one-day interactive LLM-athon where Dublin’s healthcare community puts AI to the test.
Clinicians, researchers, healthcare professionals, patient advocates, students, engineers and innovators, exploring the practical use of Large Language Models and generative AI in healthcare.
The programme combines expert keynote lectures, interactive workshops and multidisciplinary panel discussions centred on realistic clinical and non-clinical healthcare scenarios. Participants from different backgrounds will work in small groups to compare different AI models, evaluate their outputs, identify strengths and limitations, explore common failure modes, and discuss how AI can be implemented safely, responsibly and effectively across healthcare.
Whether you are completely new to LLMs or already using AI in clinical practice, education or research, the event is designed to develop practical skills, encourage collaboration across disciplines and stimulate discussion on the future of AI in Irish healthcare.
One packed day, start to finish.
Hands-on LLM activities through the morning, keynotes in the afternoon, closing with a panel on the future of AI in Irish healthcare.
- 08:30–09:00Doors Open & Check-in
- 09:00–09:15Arrival, Coffee & Settling In
- 09:15–09:45Introduction to the LLM-athonSebastián Cajas & Dr Muhammad Ali
- 09:45–10:30Activity 1: Prompt Engineering & Clinical Reasoning
- 10:30–11:00Coffee Break
- 11:00–11:45Activity 2: Evaluating LLM Responses
- 11:45–12:30Activity 3: LLM Failure Modes & Safe Implementation
- 12:30–13:30Lunch & Networking
- 13:30–14:15Group Presentations & Workshop Debrief
- 14:15–15:00Keynote: One Patient. One System. One Question: Is an LLM Ready for Grace’s Journey?Dr John Sheehan
- 15:00–15:45Keynote: AI Stands for Alea Iacta (The Die is Cast)Dr Leo Anthony Celi (online)
- 15:45–16:00Coffee Break
- 16:00–16:45Panel Discussion: The Future of AI in Irish HealthcareLeo Anthony Celi, John Sheehan, Natalie Cole, Ezi Ozoani, Noeleen McHugh, Seán Kirwan. Moderated by Muhammad Ali & Tamás Tiszai-Szűcs
- 16:45–17:00Closing Remarks
Real scenarios,leading LLMs,no spectators.
Participants work in multidisciplinary groups using leading LLMs to tackle realistic clinical and non-clinical healthcare scenarios.
The focus: prompt engineering, clinical reasoning, AI evaluation, cognitive bias, governance, ethics and safe implementation.
During the event, MedWrite.ai will provide participants with access to leading AI models, including OpenAI’s and Anthropic’s, running on servers hosted in Europe.
Prompt Engineering
Crafting effective prompts for clinical and non-clinical scenarios.
Clinical Reasoning
Testing how LLMs handle realistic healthcare cases.
AI Evaluation
Comparing leading models and critically appraising their outputs.
Cognitive Bias
Spotting failure modes in both humans and machines.
Governance & Ethics
Implementing AI safely and responsibly in healthcare.
Multidisciplinary Teams
Small groups mixing clinical, technical and patient perspectives.
AI in healthcare needs every discipline at the table.
Doctors & nurses
Clinicians of every grade and specialty who want to understand what LLMs can and cannot safely do in day-to-day patient care.
Allied health professionals
Pharmacists, physiotherapists, radiographers, dietitians and colleagues across the health service exploring how AI fits their practice.
Medical & healthcare students
Students of medicine, nursing and health sciences building the AI literacy their future careers will demand. No experience needed.
Patient advocates
Patient and public voices who will keep the discussion grounded in what matters: safe, fair and transparent care.
Researchers & data scientists
Academics and analysts working with health data who want to stress-test LLMs on realistic clinical problems.
Engineers & innovators
Software engineers, developers and founders building health AI tools who want direct feedback from the clinicians who would use them.
Healthcare managers & industry
Leaders and decision-makers responsible for how AI is procured, governed and rolled out across healthcare organisations.
No prior AI experience required. The day is designed for complete beginners and experienced users alike.
Learn from leaders in health AI.
Beth Israel Deaconess Medical Center · Harvard Medical School · MIT
“AI Stands for Alea Iacta (The Die is Cast)”
Senior Research Scientist at MIT, Associate Professor of Medicine at Harvard Medical School and Critical Care Physician at Beth Israel Deaconess Medical Center. Internationally recognised for his work in clinical data science, responsible AI, and addressing bias and inequity in healthcare AI.
Dr John Sheehan
Consultant Radiologist · Blackrock Health
“One Patient. One System. One Question: Is an LLM Ready for Grace’s Journey?”
Consultant Radiologist at Blackrock Health and Mobile Medical Diagnostics, Clinical Director of Radiology at Hermitage Clinic, Adjunct Faculty at the RCSI Graduate School of Healthcare Management and Chair of Innovation at AIMS Summit. A member of Ireland’s Digital Health Leadership Strategy Group, with particular interests in digital health, AI and emerging technologies.
Joining our panel on the future of AI in Irish healthcare.
The day closes at 16:00 with a panel bringing together perspectives from clinical practice, healthcare innovation, AI development and governance, industry, and patient and public involvement.
Dr Natalie Cole
Head of Innovation · Tallaght University Hospital
Leads Innovate Health and the development of hospital-based healthcare innovation at Tallaght University Hospital. More than 20 years’ experience across health research and innovation, including roles as Assistant Professor of Neurogenetics at Northwestern Memorial Hospital in Chicago and Change and Benefits Lead in the Office of the CEO of the HSE.
Noeleen McHugh
Patient & Public Partnership Representative · National Screening Service
Member of the National Screening Service AI and Robotic Process Automation Strategic Advisory Committee. Brings extensive experience in governance, innovation and national IT programmes, alongside the patient and public perspective on the safe and trustworthy introduction of AI into national screening services.
Seán Kirwan
Co-founder & CEO · MedWrite
Digital-health entrepreneur and former pharmaceutical executive. As Global Director of Digital Health Solutions & Strategy at Novartis, he established its Global Digital Design Centre of Excellence. At MedWrite, his work focuses on applying AI and AI-enabled workflows to reduce the administrative burden on healthcare professionals.
Joined on the panel by Dr Leo Anthony Celi (online), Dr John Sheehan and Ezi Ozoani, and moderated by Dr Muhammad Ali and Dr Tamás Tiszai-Szűcs.
The organising committee behind the day.
DubLINK is organised by a multidisciplinary team of clinicians, AI researchers and educators spanning the host venue, MIT Critical Data, MedWrite.ai and the wider Irish and international health AI community.
Senior Research Scientist at MIT, Associate Professor of Medicine at Harvard Medical School and Critical Care Physician at Beth Israel Deaconess Medical Center. Clinical Research Director at the MIT Laboratory for Computational Physiology, internationally recognised for his work in clinical data science, responsible AI, and addressing bias and inequity in healthcare AI.
Senior AI/ML Engineer with 9+ years of experience building and deploying intelligent systems. At Workday, he focuses on scaling human-AI systems. He has led and contributed to 40+ funded projects across healthcare, climate and scientific computing, collaborates with MIT Critical Data on open-source medical AI, and previously worked with CeADAR, Ireland’s National Centre for AI, and Harvard’s Visual Computing Group.
Co-founder and Head of AI at Aethon.fund and an AI safety and ethics researcher, with an MEng from Trinity College Dublin. Previously led ML ethics and red-teaming work at Hugging Face, and has worked on AI governance and safety, including with the EU AI Office on the Codes of Practice. Her work spans AI safety, interpretability, reinforcement learning and responsible AI.
Consultant Intensivist and Quality & Audit Lead in the Intensive Care Unit at Tallaght University Hospital, with an MSc in Human Factors in Patient Safety from RCSI. Leads quality improvement, patient safety and risk management initiatives, and coordinates audit and data governance. His recent work focuses on cognitive biases in clinical decision-making and human-AI interaction.
Educator and Researcher in AI and Digital Health at RCSI, where he recently completed his PhD in AI and precision medicine. With a background in bioinformatics, his expertise spans machine learning, genomics, histopathology and multi-omic data analysis. Recipient of the RCSI Vice Chancellor Innovation Award.
In the heart ofDublin healthcare.
The Venue
Workday, Building 152, 152-155 Church St, Smithfield, Dublin 7, D07 A0TN.
From O’Connell Street: Luas Red Line
Walk two minutes west to the Abbey Street Luas stop and take the Red Line towards Tallaght or Saggart. Smithfield is the third stop, roughly six minutes, and the venue is a five-minute walk from there across Smithfield Square. If you would rather walk the whole way, it is about twenty minutes along the north quays.
From Dublin Airport
Dublin Express route 782 runs to Heuston Station and stops at Arran Quay, at the foot of Church Street and a three-minute walk from the door. Coaches leave roughly every thirty minutes and take about forty minutes. Any other airport coach into the city centre also works: get off near O’Connell Street and pick up the Red Line at Abbey Street. A taxi takes twenty-five to thirty-five minutes depending on traffic.
By Car
The nearest public car park is Q-Park Smithfield on Bow Street, a two-minute walk from the venue and open from 06:30 on Saturdays. Smithfield sits inside the city’s traffic-managed core, so the Luas is usually the easier option on the day.
Made possible byour partners.
DubLINK is delivered with the support of leading Irish and international institutions in healthcare, research and AI.
Interested in partnering or sponsoring? Reach out via LinkedIn.
Thanks forjoining us.
Registration’s closed for this year. Want to collaborate on the next edition? Reach out via LinkedIn.