I am currently an M2 Artificial Intelligence student at University Paris Saclay. I worked as an NLP and Recommender Systems (RS) engineer for two years. Since it was a startup company, I had to build the entire algorithm from scratch. It includes writing requirements documents, data collection and labeling, models developing, testing, deploying, launching, and analyzing user feedback data. I enjoyed the entrepreneurial atmosphere and we were able to utilise very innovative algorithms and techniques for the product, updated and iterated fastly based on users’ feedback. That made me very excited about working as an engineer.
NLP projects mainly involve Bert pre-training model and fine-tuning, Named Entity Recognition, Multi-label Classification model, Word Segmentation, New Word Discovery, and other NLP tasks. RS projects include creating user and text-side features through the two-tower model. To improve user click-through rates, Vector search, Approximate Nearest Neighbor (ANN) algorithm, and ranking models such as Deep Interest Network (DIN) were used. Also, I was a visiting worker at the Sanger Institute, University of Cambridge. I used machine learning to predict genetic essentialism. I have experience in both the industry and academia fields, which is also my strength.