I researched on gender bias in commercials at AI4Good Lab. Under supervision of research scientists from Google Brain, Facebook AI Research and Element AI, we created a nudity detection tool with convolutional neural networks. I analyzed how females were represented in the advertising industry and discouraged gender-biased advertising techniques.
Learn more
I researched on women’s engagement in open-source software development under supervision of Dr. Laura Dabbish and Dr. Bogdan Vasilescu. I investigated potential factors that led to the underrepresentation of women's contributions to GitHub using NLP, sentiment analysis, and regression analysis.
Learn more
I worked on the TensorFlow Extended (TFX) team to democratize access to the machine learning technology through open-source software. I supported the highly demanded custom split feature for the entire TFX pipeline and researched how to incorporate Google Jax for high-performance environments. The user guide for custom splits is officially released to TFX users in October 2020.
Learn more
I worked on the Global Cyber Security (GCS) team to further their security service automation and advance their firewall for cloud operations, which is vastly consumed across RBC applications with around 10 million unit hits monthly. I created an automated service that detects and blacklists malicious URL's in emails, which protects the bank by decreasing the risk of infection and liberates 1-2 hours of manual work effort per day for the Security Operations Centre (SOC) team. I was awarded Top Accomplishment Prize by the GCS Executive and ranked 2nd overall in Fall 2020 Co-op students.
Learn more
I'm conducting HCI research on inducing empathy towards various minority groups, as well as depicting the concept of “intersectionality” through the use of VR technology under supervision of Dr. Syed Ishtiaque Ahmed. We intend on presenting our prototype and user studies in a full conference paper to CHI Play in February 2021 and open-sourcing our platform for other researchers to build on.
Learn moreI create user-centred technologies at University of Toronto's annual hackathon.
StyleList understands your fashion within a few swipes and makes personalized suggestions for your daily outfits.
I built the entire machine learning backend, which analyzes the items liked by the user and provides the next items after each swipe based on cosine similarities.
Teddie supports remote therapy for university students with anxiety and depression by being a reliable companion that students can talk to at anytime.
I built the backend for analyzing the tone of voices with natural language processing and sentiment analysis and producing the statis-tical report for mood tracker.