2020 -

Data Scientist at Standard Chartered Bank

Project: Entity Resolution on Internal to bank’s datasets and third-party datasets using streamlit, scikit-learn and Dataiku data pipeline • Developed and deployed an entity resolution Machine Learning app that identified bank customer counterparties with 95% accuracy; estimated chances of bank product sales conversion rate by 20% • Optimized natural language processing algorithms with scikit-learn cosine similarity, scikit-learn TF-IDF vectorizer and custom regex function; reduced unclean names by 40% for the search algorithm; • Implemented PySpark and scikit-learn libraries with Dataiku recipes to automate data pipeline design and development of a Streamlit app • Architected Docker images with Dockerfile and Docker compose for local and cloud deployment to provide seamless deployment

2019 - 2020

Data Scientist at HSBC Global Analytics Center

Project: Identification and Analysis of customer counterparty on trade corridors (extended Keep in House) • Built a fuzzy wuzzy algorithm in Alteryx and NLTK to identify and group 10,000+ customer counterparties in the bank's book; reduced manual effort by 75% and increased accuracy by 90% • Implemented K-Means and KNN algorithms to cluster leads into three opportunity baskets (NTB, SOW, NTT) based on revenue and product holdings; increased sales by 25% through targeted marketing campaigns (one Page PDF document) Project: Opportunity generation & Cross -Sell Opportunity for Receivable Finance & Supply Chain Solution products • Designed and deployed a machine learning-powered web application which automatically generates a single-page PDF document containing unbanked counterparties