Feb 2022 – Jan 2023 · 11 mos

Apr 2019 – Jan 2022 · 2 yrs 9 mos

Nov 2018 – Mar 2019 · 4 mos · Greater Montreal Metropolitan Area

May 2018 – Aug 2018 · 3 mos · Japan

• Researched and developed a novel semi-supervised technique to jointly train a label-enhanced
• embedding model with a highly accurate text classifier for the task of multi-domain text classification.
• Developed a variational Bayes framework to encapsulate the training algorithm which can work with
• any choice of embedding model and text classifier.
• Implemented and replicated the results of multiple baselines to compare against our model. Our
• model outperformed all baselines. Performed multiple ablation tests to identify the components
• which resulted in the performance gain.

Dec 2017 – Jan 2018 · 1 mo · Japan

• Researched and developed on applications of topic models in providing interpretable intelligent
• assistance in E-commerce domain.
• Developed a model using topic distribution vectors to evaluate and ultimately improve product
• description quality through suggestive feedbacks. Achieved 75-80% accuracy in classification task.

May 2017 – Jul 2017 · 2 mos · Gurugram, Haryana, India