Meet: Dr Georgia Richards

Today, that mission sits behind the Preventable Deaths Tracker, an initiative bringing together data from coroners’ Prevention of Future Death reports to uncover patterns, improve transparency and identify opportunities to prevent similar deaths from happening again. Georgia’s work spans health, justice, informatics and policy, using methods more commonly associated with epidemiology to draw insights from the coronial system. 

Georgia Richards is an AI+ Senior Fellow at King’s College London. These interdisciplinary fellowships are central to King’s £18 million strategic investment in academic excellence and demonstrate King's commitment to transforming AI research and innovation across all disciplines.

Georgia’s route into this unusual intersection began in Australia. While studying pharmacology and working part-time in a pharmacy, she became interested in why people were taking pain medications for longer than what was prescribed. An honours research project in a multidisciplinary pain clinic deepened that interest, particularly the disparity she saw between research evidence and what was happening in clinical practice.

It eventually brought her to Oxford’s Centre for Evidence-Based Medicine, where her PhD investigated the use and long-term harms of pain medications. Alongside traditional sources such as NHS prescribing data, Georgia became interested in harder-to-reach sources, including private and over-the-counter opioid use.

"There’s so much that we don’t know," she says. "We can’t really understand the full extent of a problem if we don’t have the right data."

When she began exploring how opioid-related deaths might be prevented, Georgia discovered another overlooked source: information from coroners and death investigations. Unlike Australia, where a national database of coronial inquests existed, she found there was no equivalent resource bringing this information together in the UK.

The solution began unexpectedly.

"It started as an accidental coffee conversation," Georgia recalls. A fellow PhD student offered to help her create a simple web scraper to collect Prevention of Future Death reports from a government website. What began as code on GitHub gradually developed into the Preventable Deaths Tracker.

During the COVID-19 pandemic, its wider potential became particularly apparent. While researching opioid deaths, Georgia began noticing repeated deaths involving alcohol-based hand sanitiser. She analysed poisoning data and identified groups that could be at particular risk, including children and young people, people with mental health vulnerabilities, and people with dementia living in care settings. and people with dementia living in care settings.

It also convinced her that the data needed to become accessible beyond researchers who knew how to code. "I can't just have a piece of code on a GitHub account," she remembers thinking.

The path since has not been straightforward. After struggling to secure academic funding for work that did not fit neatly into conventional definitions of healthcare research, Georgia left academia while continuing the Tracker as a passion project. She later used money earned working as a safety epidemiologist for AstraZeneca to fund its development, before returning to academia and joining King’s as a King’s Prize Fellow in June 2024. It was the first time she was funded by a university to work on the Tracker as research rather than alongside another job.

At King’s, the project has entered a new phase. The Tracker has evolved into a more sophisticated case management system, with automated collection, cleaning, and processing of data. Georgia is now beginning to explore how generative and non-generative AI models can accelerate that work and identify insights that might otherwise remain hidden.

One priority is making thousands of digital files easier to search and analyse. New tools can help identify relevant records in seconds, after which Georgia and her team apply their expertise to verify and interpret the results. They are also exploring AI-assisted data extraction and whether models can reveal new trends and insights across the collection. 

This combination of data science, healthcare and law has placed Georgia in an unusual disciplinary position. Having worked extensively with coroners and developed expertise around coronial records, she describes her work as bridging health and justice. 

"I’m often hearing people telling me what I’m not," she says. "And now I’m trying to help my students say what we are. We are methodologists, we are epidemiologists, we do have a set of important skills."

She sometimes describes herself more simply as a "data detective", pulling apart information to ask whether the conclusions being drawn from it are really true and to make the processes behind those conclusions more transparent. Her work has convinced her that even something that appears as straightforward as how a death is recorded can have far-reaching implications for national statistics, policy, funding and public services.

Behind the data, however, are people. Georgia says contact with bereaved families is a major reason she continues to work in an area she acknowledges can be difficult. Families have also become important partners in her policy and parliamentary work, helping bring evidence about preventable deaths directly to decision-makers. 

For Georgia, the interdisciplinary nature of the AI+ Fellowship feels particularly appropriate and it has given her the opportunity to devote herself to research that began almost by accident.