Confidential, New York, NY
Chief Data Scientist
- Working for an AI/Data Science consulting firm serving clients in finance, insurance and healthcare sectors.
- Evaluate the client’s data and analytics capabilities given their goals and mission, and recommend advanced analytics infra/methodologies, considering their historical performance and practical constraints.
- Create POCs relevant to the client’s cases to provide them with a strong sense of feasibility on projects or proposals, involving Deep Learning, Natural Language Processing, Chatbots, Machine Translation, Image Processing, Speech Recognition and others.
- Keep track of latest AI/Deep Learning trends and actively be engaged in communicating with senior data scientists across industry through conferences to gain new analytics insights.
Confidential, Wilton, CT
- Developed several retail credit models covering mortgage, credit cards, auto loans, small business loans using regression, time - series and survival analysis with SAS.
- Developed alternative models in Machine Learning, and AI/Deep Learning frameworks for internal validation and research purposes using Python, TensorFlow, Keras and PyTorch.
- Developing Natural Language Processing and Recommendation System based models targeted at understanding customer default, pre-payment, and cross-selling patterns across segments and time dimensions—all back ended by Deep Learning (CNN, RNN, LSTM and GRU) using Python, TensorFlow, Keras and PyTorch.
- Authored white papers on modelling practices with stress-test results and submitted the documents to FRB and OCC for formal reviews on the bank
Confidential, New York, New York
Chief Analytics Officer (CAO) and Senior Data Scientist
- Built up analytics infra for a start-up hedge fund specialized in peer-to-peer lending from day one and critically contributed to the firm’s AUM growth to more than $100M within three years from zero-funding at the beginning.
- Led a team of scientists and developed credit risk and revenue forecasting models, and used regression, survival analysis, decision tree, artificial neural networks, SVM, and others and created Ensemble Modeling, Random Forest, Gradient Boosting based models to deliver optimal analytics.
- Enhanced the model’s predictive power, utilizing results from Sentimental Analysis, Topic Modeling, and other Natural Language methodologies.
- Utilized Hadoop and Spark framework for effective data access, storage and analytics processing in memory to further process the data in larger quantities.
- Converted some of Machine Learning based applications to Deep Learning applications (CNN and RNN) using Python, TensorFlow and Keras.
Confidential, Long Island City, New York
Applications Development Senior Programmer Analyst
- Implemented PD/LGD/EAD models to meet Basel II and III and compliance requirements using SAS, SAS Enterprise Miner, VBA, and R.
- Developed credit risk/market risk/counterparty credit risk analytic models using SAS, SQL, and R for analytics involving mortgage, credit cards, and secured and unsecured loans, maturations, vintages, and macroeconomic variables.
- Designed collaborative filtering (a recommendation system) using k-nearest neighbors involving cross-selling of various financial products based on customers behavior and risk profiles.