Helped bring a Machine Learning solution to production to increase the rate of processing of Veteran's claims for Toxic Chemical Exposure using AWS Comprehend, AWS Sagemaker studio, Docker, Jenkins, Harness as a subcontractor to Booz Allen Hamilton on the BID project

• Technologies: Python, Docker, Grafana, SQL.

Software Engineer II - SCOUT Ad Hoc | January 2022- Present

• Created a dashboard that allows administrators to add new users to FISMA systems using React, Typescript and GraphQL

• Utilized Ruby to build new a new module to record information about relevant AWS Lambda services for the System Inventory Tool (SIT)

• Upgraded the SCOUT project through a large breaking version change including from React 17 to React 18 and from React Router 5 to React Router 6

• Used Cypress to create integration tests for the SCOUT inherited data

• Fixed a bug so that the focus was correctly set to the artifact table after uploading a file

• Created mock data and updated the inheritable job so that it was pulling the mock data from s3 and removed all the ARS 3.1 logic

Technologies: React, Typescript, GraphQL, Cypress, Jest, Docker, Ruby

Software Engineer II - MAPI Ad Hoc | February 2019 - January 2022

• Responsible for updating the Marketplace API and Find Local Help API as well as the coverage portal

• Added tooling to manage GovDelivery emails for the provider coverage for MAPI

• Created the ICHRA endpoint for MAPI

• Updated all the relevant calculations for the post 2022 minimum tobacco use age to set it at 21 years for MAPI

• Created a basic Postman collection for new users for FLH API

• Conducted Jmeter testing which determined that the MAPI could handle an estimated increase in traffic during the open enrollment period

• Added a 15-minute idle timeout for a session token for FLH API

Technologies: Golang, Python, React, Bash, Node JS, Docker

• Portfolio Optimization – Identified the profitability of a high risk / marginal segment and executed solutions to decrease risk and

improve portfolio. Used python to clean and organize data, used machine learning framework to evaluate / select optimal model.

• Production Solutions – Tackled deployment issues by creating an executable model derived from an abstract model. Used cutting

edge synthetic data creation techniques to train a Neural Network with limited original data.

• Business Creation – Created Monte Carlo simulation on disapproved applicants to identify business opportunities and cost saving

strategies with high risk applicants.

• High Fraud­Risk Deal Evaluation ­ Developed several strategies to eliminate deals that had a high risk of being fraudulent.

Wrote code in python to collect tweets from the twitter API then do natural language processing with the TextBlob package based on NLTK and determine the subjectivity and polarity of each tweet.

Generated topics using Latent Semantic Analysis and Latent Dirichlet Allocation from the collected tweets.

Processed simulated catalogs created by the Dark Energy Survey using python to select objects behind simulated dark matter clusters to measure the shear profile with true and estimated photometric redshifts.