Narrowing down Michelle Lee’s list of accomplishments and accolades into just five highlights is a fool’s errand. On paper, Michelle’s career and academic pursuits seem to tell you everything you need to know—based in London, she’s the senior manager and artificial intelligence (AI) ethics lead in Risk Analytics for Deloitte UK and a Ph.D. candidate in computer science and technology at the University of Cambridge. Thanks in large part to Deloitte’s flexible working arrangements, Michelle works one day a week with her Deloitte team, developing a suite of offerings focussed on building trustworthy AI and advising clients across industries on how to manage new risks introduced by AI. The remaining six days of the week are dedicated to her Ph.D. pursuits. And this doesn’t even begin to scratch the surface of Michelle’s success in the risk and analytics industry.
As Deloitte honours exceptional and inspiring “Women in Risk Advisory”, we’re shining a spotlight on Michelle, her career, her accomplishments, and more. Read on for five facts about Michelle Lee:
After graduating from Stanford University with her bachelor's degree in 2014, Michelle joined Deloitte in a strategy and operations consulting role to explore diverse industries, functions and domain areas. In 2015, when she decided to move to London, a partner she was working for at the time recommended her to Deloitte’s AI and Data Science team and it happened to be within the Risk Analytics function. Michelle jumped at the opportunity and has been with the team ever since.
Over the last seven years, Michelle has led the design and development of a suite of offerings and technical accelerators for building trustworthy AI. “It’s exciting to be at the forefront of this constantly changing landscape that is AI and risk,” Michelle shared. “I’ve loved the constant push for innovation.”
During her time at Deloitte, Michelle has worked on a number of AI tools, including building a chatbot for risk teams, called EMILIE, which is now being used across Deloitte North-South Europe; developing a geospatial neural network model to identify high-potential investment locations for an ATM service provider; and creating a scheduling optimisation tool for a hospital. But Michelle’s career has recently focussed on helping organisations think about the regulatory and reputational risks associated with AI systems, as well as the unintended societal impact that can lead to unethical consequences when AI is not built properly.
“AI learns from the past and human behaviour, and that includes all of our own biases, so unless we adjust for them, AI will perpetuate these stereotypes and exacerbate any inequalities that already exist,” Michelle told us.
A big part of what Michelle does at Deloitte is focussed on building the Trustworthy AI offerings-especially fairness, explainability and robustness tools—to help organisations make sure the technology that’s being built represents and works for everyone. She recently led the development of Deloitte’s AI governance framework and controls library, which is accessible across risk domains to help organisations accelerate gap analysis as they scale and productionise their AI systems.
“There are algorithms being built across industries that inform decision-making that affects a lot of people. Everything from insurance pricing, credit risk evaluation, fraud detection, recruitment and hiring. Organisations should consider a more systematic approach for managing the risks brought about by these systems,” Michelle said.
While building ethical AI practices for her clients takes up most of her day, Michelle is passionate about using AI to achieve positive societal impact. She’s currently a member of the Ethics Committee for a pro bono data science organisation and she has worked on several projects to help a variety of non-profit organisations leverage AI and data to make more informed decisions about their work.
At the pro bono data science organisation, Michelle works with data scientists and helps run the training sessions on how to ask the right kinds of ethical questions when developing AI systems. She leads events for the organisation, and she even fundraised for a DataDive (a data science hackathon) to help a cancer charity understand bone marrow and blood donor attrition to inform their retention strategy.
At Deloitte, Michelle worked with a Deloitte’s partner charity tackling UK youth homelessness to use data to identify areas in high excess demand of homeless shelters. The project won a DatSci award in 2018 for Best Use of Data to Achieve Social Impact and was also featured as one of Deloitte’s best achievements in 2018 in the Deloitte Impact Report.
In 2018, Michelle decided to go back into academia. She was in the process of building out the first version of the Deloitte fairness toolkit when she realised a lot of work coming out of academia did not align with real-life applications. “I started writing about how to think about fairness and ethics in more practical, real-life settings and I ultimately decided to pursue a Master of Science at the Oxford Internet Institute,” Michelle explained. It was at this time that she moved to part-time work at Deloitte.
And she didn’t stop there. In 2019, she began to pursue her Ph.D. at the University of Cambridge in computer science and technology.
“Without the flexible working scheme at Deloitte, I wouldn’t have been able to pursue this work at all,” Michelle said. The flexible working arrangements enable employees at Deloitte to temporarily change their working hours with partner approval.
“There is a lot of synergy between my work and my research on AI risk and ethics—one informs the other,” she said. “My academic career and all of my research have made me a better subject-matter expert and a better adviser for clients. My work experience has made me a better researcher because everything that I publish is founded in practical challenges that I’ve seen clients face before.”
Currently, Michelle is in the process of completing her thesis and hopes to submit it in September 2022, a short three years since starting the degree.
Michelle said her career has not been perfect, but she’s remained steady through the ups and downs, thanks in large part to her mentors. As a woman working in several traditionally male-dominated industries—risk, financial services, and technology—Michelle noted that she has sometimes faced misperceptions.
“I was often underestimated because I look very young, as I’ve been told, and I’ve gotten comments from people that, ‘You don’t look very technical,’ or ‘You’re not really what I expected of a computer scientist,’” Michelle said. “I guess there is a certain image or stereotype that people have of individuals working in AI, especially. But I’ve been fortunate to have champions and mentors in my career who are really invested in my growth and in my development, who have challenged me and advised me.”
And while Michelle is very passionate about her work, she balances her day-to-day work with a variety of activities. She’s currently a member of the rowing team at St John’s College at Cambridge and a member of the university’s Blind Wine Tasting Team, which recently won the Varsity Match against Oxford. She’s also a runner and participates in the Medoc Marathon in Bordeaux annually.
View Deloitte’s Women in Risk Advisory series