▸ Built and deployed a large-scale entity resolution system consisting of a content-based recommender engine and Louvain algorithm for graph partition using Spark, Scikit-Learn and GraphFrames, leveraged NLP techniques such as word embeddings and named entity recognition (NER) in feature engineering💪 Impact: consolidated over 300K customer accounts across 3 ERP systems; realized >$100M shift in customer opportunity understanding; enabled B2B e-commerce▸ Trained and deployed a tree-based classifier to predict regular and irregular demand using Spark ML, HyperOpt and MLflow, model served as a REST endpoint with Azure App Service💪 Impact: increased prediction accuracy and reduced annual overstock▸ Built and deployed a full-stack web app recommending optimal allocation of sales force using rule-based logic from statistical distributions of customer attributes and approximation of geolocations💪 Impact: achieved 90% efficiency gains in annual planning for 9 siloed regions of sales leadership; increased 22% YoY revenue for early adopting regions▸ Initiated and led team(of 9)'s migration to Azure and adopting git version control, code review, Agile sprint planning with Jira, Spark programming, automated testing with Pytest and Cypress, and CI/CD with Azure Pipeline💪 Impact: resolved ML model lab-to-production challenge; significantly increased team efficiency and codebase quality

▸ Implemented linear regression model to predict ad performance across multiple channels using R💪 Impact: increased digital revenue by 44% and saved $11K marketing budget▸ Identified retail expansion opportunities through mining large amount of semi-structured and unstructured data in Salesforce and Google Analytics using SQL and R💪 Impact: launched 3 new stores and generated $800K revenue in first three months of opening▸ Provided insights on customer segmentation, A/B testing and campaign structure for email and SEM channels💪 Impact: improved YoY conversion rate by 29%

INVISTA is one of the world's largest petrochemical companies and part of the multinational giant Koch Industries. I was an analyst in a newly-formed strategy team of 5 and developed commercial strategies for the then-under-construction plants worth over $500 million. My key contributions include:▸Developed a market-based pricing model by modelling supply and demand using polynomial regression💪 Impact: reduced trading risk by 18% ($90M impact annually)▸ Initiated and led database system automation tasks using SQL💪 Impact: saved 90% of monthly updating time from one week to 1/2 day