NLP at Bloomberg AI

• Recommendation & classification
• NER/NED/Salience
• Topic classification
• Human-AI collaboration for continual training of production NLP models
• Sequence tagging, slot-filling, structure extraction from texts
• Pretraining and fine-tuning of LLMs

Recommender Systems at GroupLens

• A generic recommender and predictor server.
• GitHub Link: https://github.com/grouplens/samantha
• Website Link: https://grouplens.github.io/samantha

Other Projects

• Zero-inflated Regression Problem, Prediction Uncertainty Estimation and Scaling Up of Bootstrapped Models (Wise.io)
• Sequential query pattern mining in Hadoop, distributed training of matrix factorization models in MPI, feature engineering in Gradient Boosted Decision Tree (Baidu Inc.)
• Research and development of the video recommendation systems, user preference modeling and video topic modeling (Funshion Ltd.)