My Projects
Within this portfolio, I highlight a few key projects across my career. Projects that have access to code and data have their respective GitHub links attached. If you would like to learn more, please do not hesitate to contact me.
Workforce Wellbeing
To learn more about this work, please contact me through this form.
In my work as a research scientist, I had the opportunity to lead a research team on client-sponsored work.
The problem: A client wanted to understand how their workforce's wellbeing and performance was impacted by the kind of work they do.
The approach: Mixed-methods. In my role, I led two research streams: a quantitative stream investigating the impact of shift on performance and wellbeing, and a qualitative stream creating custom surveys and interviewing employees to understand general impacts on wellbeing at work.
The results: Our quantitative work revealed that key KPIs were impacted by shift, and we were able to quantify that impact and provide our client with recommendations on how to bolster performance while maintaining wellbeing, with respect to shift. Our qualitative work uncovered employee sentiments that were impacted by their work, which were delivered to the client alongside research-backed recommendations for maintaining and bolstering morale customized for their workforce.
Brain Gym
Learn more about BrainGym from Technology Lead Tayfun Esenkaya during an interview with MongoDB.
One of my priorities at Accenture as a Research Scientist was utilizing neurotechnology and AI to transform the way we approach wellbeing.
The problem: Workforces are under increasing pressure to deliver value to their company and the company's stakeholders, and that pressure may increase stress and decrease resilience.
The approach: We partnered with neurotech startup Mendi and leveraged MongoDB AI capabilities to provide individual and workforce-level insights into wellbeing and resiliency, as measured by bio sensing technologies (Apple Watch, FitBit, Oura Ring, etc.) and work-related productivity data (emails, tasks, notifications, etc.).
The results: We created a proof-of-concept platform called BrainGym that integrates these data sources and provides individual-level and workforce-level feedback related to wellbeing and productivity. This platform also includes an AI chat assistant that integrates historical data and insights from neuroscience & wellbeing to provide personalized feedback. BrainGym debuted at CES 2024 and was awarded Billboard's "Best Brain Hack" of 2024.
Learning and Training
In my work as a research scientist, our group was interested in exploring how AI assistants can work alongside humans in their work, and train people how to do physical tasks efficiently. My role on this project was to conduct a user study and understand how people would respond to an AI training coach.
The problem: Are AI assistants more effective at training than traditional training methods?
The approach: A Wizard-of-Oz study, where I built and controlled a simulated version of an AI-assistant, and conducted a 2x2 experiment, teaching novices how to make two different complex coffee drinks with the help of either the simulated AI-assistant, or going through a training video at their own pace. We collected information on the amount of time the training took, as well as user perceptions around both training experiences.
The results: We found that while both training experiences took approximately the same time across all users, the AI assistant that guided people through training was found to be instill more confidence in users, as well as perceived to be more effective.
Read more about this user research study here on Medium.
Bimanual Coordination
In one of my projects in graduate school, I investigated how our arms synchronize during a reaching coordination task. In this experiment, we wanted to understand how haptic manipulation - through applying forces to the hands - influenced how our arms are coordinated. Check out the video on the left to see what we found!
​
​
If you'd like to see the code and data I used in this project, check out my Github page.
Meta Analysis on Brain Stimulation
As a graduate student, I had the opportunity to work on an impactful project assessing how effective facilitatory brain stimulation was on human cognition. By stimulating the brain with magnetic fields in bursts with particular timing, you can actually prime neurons for activation and make them more likely to fire!
The problem: Many researchers are using facilitatory brain stimulation on different brain areas with different techniques, and in the state of the field, we had no idea what works and what doesn't.
The approach: Using meta-analysis and systematic review, we empirically assessed whether or not this kind of stimulation impacted healthy human brains and facilitated cognitive processes (like memory, attention, motor skills, etc.).
The results: We found that when accounting for outliers in the data, we did see a small significant facilitatory effect - but much of the variance of the effect was caused by researchers not reporting the full scope of stimulation parameters - so we came up with a set of guidelines for researchers to help assess this problem in the future.
