Citibank

Sovereign Overnight Financing Rate (SOFR)

For Citibank, we supported transition the transition from using the London Interbank Offer Rate (LIBOR) as a daily target interest rate to SOFR. Developed finance requirements for Next and Prior Reset Rate Dates calculations for Interest Rate Swap products to ensure daily risk modeling and calculation capabilities following LIBOR cessation. Identified downstream system impacts, quantified the impacted trade population, and coordinated with multiple finance, data owners, quantitative developers, and technology teams to prioritize, schedule, and complete end-to-end testing prior to LIBOR cessation date.

The Gates Foundation

Starism

As a contractor to the Bill & Melinda Gates Foundation’s Institute for Disease Modeling, we advanced Starsim; a stochastic, agent-based disease simulation modeling tool that was open-source, written in Python, and not ready for public release.

We reviewed previous IDM tools, analyzed the Starsim codebase against Python best practices, reviewed user stories to develop a voice of the customer, performed time and memory profiling, and we prioritized Github Issues based on critical-to-quality features. We created and implemented a go-to-market plan that leveraged review insights and identified and prioritized critical-to-quality features.

Reduced MVP public launch time from 6 to 4 months—a 33% reduction in cycle time.

US Department of Health & Human Services

Regulatory Report Automation

Developed a monthly report for mammography facilities was manual, time-intensive and involved data acquisition from various data stores.

We developed data pipelines in an AWS cloud-based environment triggered with AWS Lambda event-handling and automated the report using Python.

Reduced monthly reporting of regulatory metrics from 5 days to 2 minutes–a 99% reduction in reporting time.

US Center for Medicare & Medicaid Services (CMS)

Evaluation of Dialysis Facilities

Refactored Apache Spark Scala code to eliminate unused classess and methods, reduce cyclomatic complexity, and enable Scala methods to align with method lengh scalastyle requirements resulting in improved readability, maintainability, and a reduction in over 2500+ lines of code.

Transformed legacy Scala code to Python to automate validation via AWS Lambda and pull data sourced from Amazon S3 buckets for dialysis facilities’ calculated effectiveness scores against an independent consulting firm for the annual quality incentive reporting process.

Happy Money