• Primary Developer on the cloud-based mortgage pricing engine behind Mastr, an AWS Cloud-based mortgage analytics API with web and Excel-based front ends.• Optimized prepayment model and generation of mortgage rates, including a 3x improvement in the speed of the spline-interpolation code.• Improved the efficiency of data caching in the cloud, including building a custom Redis client library.• Optimized the distributed pricing engine, where packets of simulation paths are run concurrently on different machines, including the use of a continuously compounded spread so that cashflows can be aggregated first, and price to OAS computed later.• Added support for Term Sofr indexes (Libor replacement)• Investigated and corrected critical program crashes and memory leaks

• Translated research models of prepayment, delinquency, default, loss severity, house price appreciation, counter-party risk, and mortgage insurance from R, Matlab, SAS and Excel into production models in C++.• Wrote a customizable logistic model for rapid development and testing of research models.• Worked closely with quantitative researchers and trading desks to add capabilities, investigate failures, and resolve issues.

• Lead developer on PLEX (PnL Explained), a new tool attributing changes in market price to root causes (interest rates, prepayments, amortizations).• Created daily risk metrics for Securitised Products Group for regulatory reporting.

• Developer on the C++ application framework that simulated agency and non-agency cash flows, behavior models, and fixed-income analytics as member of six-member team.• Performed major overhaul of the framework to replace a critical legacy system that calculated the cost of defaults in the Freddie Mac agency mortgage portfolio.• Expanded the analytics capability of the framework so that it could run different analytics at different levels of aggregation while still providing the minimum number of cash flows required for later stages of aggregation.• Created a presentation on the new capabilities of the framework that was so well received it is still used for training purposes today.• Modified existing behavior models and added new models into the framework.

• Team Lead for Bond Agent, a bond waterfall simulator and pricing-model of non-agency CMOs (collateralized mortgage obligations) in C++ that is still a key part of the MIAC Analytics software suite.• Overhauled LIBOR Market Model (LMM) in C++ to run 10x faster to meet client performance expectations.• Developed Key Rate Sensitivity Tool, a tool for calculating partial durations (KRDs), in C++.• Key programmer for VAST, a tool to test the sensitivity of a portfolio of mortgages or MSRs to various shocks in C++. • Improved accuracy of the yield-curve bootstrapping and efficiency of the Monte Carlo rate-generation in C++.• Led projects for JPMorgan Chase, Citibank, Washington Mutual, ABN AMRO and IndyMac.