• Simulated general-equilibrium models on MATLAB to estimate the effect of central bank discount window policies on inflation. • Collected datasets from public sources (Census, FFIEC, Fed, BLS), cleaned and analyzed using R.• Employed cutting-edge causality regressions: instrumental variables, doubly-robust estimators, differences-in-differences, and event studies to analyze effect of fiscal policy on bank lending.• Constructed behavioral models to show optimal actions for banks under aggregate liquidity risk.

• Estimated the causal impact of Fulfilled by Amazon for sellers using synthetic control and conducted robustness and sensitivity analysis using machine learning (random forest, neural net).• Collect big data (10 mil+) from AWS S3 using SQL and performed analysis using R and Python (Sagemaker) to implement proof-of-concept for Amazon’s SellerCentral.• Wrote and presented business and technical documents in OP1 for executives and directors. Analyzed and summarized key figures using Python (matplotlib, pandas).• Assisted team with estimating the causal impact of account managers for cost benefit analysis. Showed that there is heterogenous effect of some AMs, allowing for optimization.

• Be the point of contact with 5 data vendors to negotiate pricing, set up test batches and receive and implement returned data for company to make better informed decisions• Create Excel modules with formulas and VBA for different departments to improve efficiency and automate manual entry tasks• Assist the setup of SQL to Access database and use PowerQuery and VBA to merge XML, CSV, and JSON formatted data from data vendors into inventory• Wrote VBA code to automate sending payoff and reinstatement requests, loan analysis, payoffs and reinstatements to save hundreds of manhours per year• Created a new standardized template for analysis of financial models using credit profile of borrower and market trends within the first month• Pulled data from open source APIs to implement into our existing database using VBA.• Developed risk models using a logistic approach to increase loan repayment probability by 40%.