Data Scientist Leading the Way with Machine Learning

Mitchell McCutcheon is Ryman Healthcare’s first Data Scientist, working on Ryman Healthcare’s innovative AI machine learning journey. 

“Ryman has a massive amount of data and we want to use that to the best of our ability,” said Mitchell.

Mitchell’s role is to use this data to project forward by creating predictive and forward-facing machine learning models. 

For his first major Ryman project, Mitchell built a falls-prediction machine-learning model that uses 135 points of data to gain insight into trends and can assess 3,500 residents in five minutes, the same amount of time it takes to assess one resident manually using a six-point questionnaire. 

To complete 3,500 resident falls assessments via this six-question survey would take around 300 hours, but Mitchell’s model can do all 3,500 assessments in the time it takes to manually complete one. Using this model, millions of assessments could be done, 24 hours a day, seven days a week.

The model interlinks a lot of different criteria, such as whether a resident has fallen in the last 30 days, the number of days since their last fall, or whether a resident is sight-impaired, to create a falls prediction rating for each resident.

“It is very complicated because if predicting falls was easy, we’d all be doing it. We’ve had some fascinating and promising results so far, so we are very optimistic that in future we will be able to advance this particular cause through the use of AI, work that Mitchell has led the way for with Ryman,” says Rick Davies, Head of Technology and Innovation at Ryman.

“The model would say, what’s this resident’s risk level, and what is the likelihood of them falling within the next seven days,” explains Mitchell.

“It’s utilising the advanced computational capabilities of computers and vast amounts of data to mathematically approximate and model a particular pattern, behaviour or phenomenon present in the data.”

Mitchell’s Ryman journey

“Mitchell joined us as a super-smart graduate,” said Rick Davies.

"He gets to lead conversations with world-class AI partners that Ryman works with and gets to learn from them as well.”

Prior to joining Ryman, Mitchell completed a degree in Accounting, Finance, and Data Science at Otago University. When COVID-19 hit and job opportunities dried up overnight, his aunt who was a Village Manager at a Ryman village let him know that temporary Security Guard jobs were available at the village gates.

While being interviewed for a security job, Mitchell was asked what he would like to go beyond that and mentioned his background. Ryman recognised Mitchell’s skills and potential and ultimately created the Graduate Data Scientist role within the BI Reporting team specifically for him, and he has now been promoted into a Data Scientist role.

Innovation at the lunch table

Mitchell describes Ryman as having a culture of innovation, even at the lunch table. He can access platforms that he wants to use and employ global thinking in the holistic sense around the health and wellbeing of the residents.

Mitchell is looking forward to being more and more innovative as Ryman grows and continues to explore cutting edge technology. Because technology is always evolving, external developments in the tech field are exciting too. 

“The culture here is great, it encourages me to try new things, we gather ideas with thoughts of projects to come, and Ryman is very supportive of continuous learning,” he said.

“You can be as crazy and innovative as you want because even if we can’t do it now, we could do it in the future.”

A multi-faceted journey

Mitchell can apply data science to any facet of Ryman’s journey and has a long list of potential projects in the pipeline, including a construction project which would predict the length of the development of a new village, informing on factors such as health and safety, cost prediction, and which teams would work best together. 

Another potential use is recommending residents meals they might like, or activities they might enjoy, or predicting where residents would want to live and how they want to live. 

Ethics and privacy are always important considerations with all data involving residents. 

“Data science tells you what information matters, and often that will surprise you. There can be lots of data out there that you wouldn’t necessarily associate with effects,” said Davies.

“Any given prediction is going to rely on the bits of data that influence the outcome, we don’t necessarily know what that will be, but that’s the beauty of Mitchell’s work.”