Addison Putnam, Ph.D. Email & Phone Number

Salem, Oregon, United States

Remote

• Utilized Natural Language Processing (NLP) and LLMs to create rich text features and incorporate large text data into modeling workflow through word embeddings.• Built Deep Learning Time Series models to forecast future stock market trends.• Applied Topological Data Analysis (TDA) on financial market data to prospect options inline with current risk and modeling focus and improve marketing capabilities.

United States

• Utilized and automated Natural Language Processing (NLP) techniques on unstructured text data from over 20 million websites to create word embeddings uniquely tailored to each project.• Applied linear algebra and vectorization to decrease runtimes of commonly used Python scripts by up to 10 fold.• Automated and standardized the data ingestion, cleaning and results visualization processes reducing time to complete projects by over 1/3.• Collaborated with data engineers to source data, independently completing every stage of model development including experimentation, training, testing, deployment to production and monitoring for each project.• Created a dimensionless visualization summarizing the shape of our data lake containing over 100 million companies and hundreds of different metrics using Topological Data Analysis and taught the sales team the needed theory so they could incorporate this into their presentations.• Interacted with multiple branches of the company including sales, marketing, customer experience and the leadership team to problem solve, design experiments and leverage my skills to support the goal of automating the underwriting experience for P{\&}C insurance companies.• Enhanced the value of existing models by increasing target space by 3 fold without losing model performance, documented models in detail and presented results to stakeholders.

Corvallis, Oregon Area

• Completed multiple modeling projects in applied statistics courses using R, including a study on the gender pay gap by using linear regression to account for non-discriminatory variables for starting pay and then accurately assessing if there is a significant difference in pay across genders.• Researched theory of and best practice techniques for linear and generalized linear modeling, including how to deal with censored data and other challenging data issues.• Planned and delivered engaging lectures in probability and statistical modeling and provided personalized tutoring.

• Created time series forecasting models of exchange rates and learned best practices for modeling financial data.• Applied extreme value limit theorems including the Generalized Pareto Distribution for modeling rare events to better predict sudden jumps in financial data.• Built copulas to measure performance of different distributions on the tail distributions of the given financial data.

• Cleaned and transformed data and developed pricing model for Personal Auto insurance by utilizing current actuarial methods on 200,000+ rows of data.• Created an Rshiny application to automate the process of model fitting, data visualization, and reporting model performance.

Doctor Of Philosophy - Phd, Mathematics And Statistics

Master Of Science - Ms, Mathematics

Bachelor Of Science - Bs, Mathematics

Activities and Societies: Pi Mu Epsilon member