Nicholas Higgins Email and Phone Number

Model Development: Python implementation of an asset back security loss model utilizing OLS regression. Responsibilities inclusive but not limited to data procurement through both Bloomberg and internal oracle databases, model prototyping, model development and integration into an existing framework, and documentation. Implementation of said model freed up 19 million in collateral.Productionalization: Productionalization of asset backed models onto a general computing platform built on PySpark. Implementation in PySpark cut run times down from hours/days to minutes.Model Framework: Implementation and Design of a framework for python models. Design was the largest part of this project as it involved identifying what components models shared and writing common interfaces and compound Objects. (insert code pathway cutdown)Compound model Framework and Runner: Design and implementation of a framework that allows for the execution and aggregation of multiple model results. The benefits of this is that sub models can be designed and implemented independently while still being easily incorporated into the final results.

Model Development: Utilization of a deep neural network and fraud data to develop a model that matched the industry standard for detection rate.Document Automation: Development of a tool in python to connect to an internal database and automatically prepopulate both Word and Latex Documents.

Web Tool Development: Use of HTML5, Bootstrap, and PHP to develop web tools to search and interact with internal databases. The pages interacted with both Oracle MYSQL and IBM Domino databases.

Automation Scripting: Utilization of Ruby to connect to an array of Linux servers, run a series of Bash commands, and compile the output into easily readable csv files.