Aaron Brake Email & Phone Number

Promoted to Mid-Level Data Scientist at AllstateModel Creation Leadership:- Collaborating with the in-house labeling team for streamlined data flow- Taking charge of other team members' models, demonstrating expertise of model creation and improvementEnd-to-End Model Development- Creating three models from inception to deployment in a pipeline- Utilizing internal modeling processes for comprehensive model developmentContinuous Learning and Impact:- Seizing opportunities for professional growth and learning- Committed to driving impactful results through innovative model creationExpanded Autonomy and Work Direction:- Independently crafting agile stories and providing strategic direction for projects- Enhanced control over work direction and project implementationCollaborative Future:- Eager to continue contributing to Allstate's data science initiatives- Excited about the journey ahead and driving the team to new heights of success

Examined and documented text sentiments in unstructured message data utilizing SBERT clustering techniques and presented findings to business partners that resulted in project approval for further exploration and modeling. Collected and cleaned chat message using PySpark and web traffic data for modeling purposes. Employed data mining techniques to extract meaningful insights and transform text into actionable data fields, improving the accuracy and efficiency of downstream analyses and modeling. Managed and directed a team of data scientists and business partners in labeling text data, resulting in the creation of supervised models that offer valuable insights to decision-makers. Developed multiple text sentiment classification models utilizing XGBoost and BERT, contributing to the enhancement of the chatbot framework. Demonstrated an ability to stay up to date with cutting-edge techniques and collaborate effectively with other teams. Programmed a script that consolidates multiple predictive models and pipelines, providing relevant statistics and metrics to support business decisions and guide future projects. Exhibited strong programming and analytical skills, and demonstrated an ability to create tools that facilitate effective decision-making. Analyzed text sentiment model results which uncovered key insights that improved the overall chat experience. Identified common reasons for customer callbacks and roadblocks faced during chats, presented findings to business partners which drove strategic decisions to enhance customer satisfaction.

Raleigh, North Carolina, United States

- Communicated statistical ideas and concepts to undergraduate and graduate students through teaching classes and holding office hours-Coordinated with other TA’s and professors to manage course logistics and develop assignmentsTeaching Experience:Python Class TA: Guided undergraduate students through their first programming class, using Python.Graduate Statistics TA: Assisted non-statistics majors in mastering statistical concepts.Undergraduate Statistics TA: Supported undergraduate students in understanding statistical principles.

Morgantown, West Virginia, United States

Classified the alumni most likely to join the Association using an Alumni Satisfaction Score metric which increased direct mail campaign response rate by 50%. Collaborated with Media, Social Media, and Graphic Design teams to analyze the success of email campaigns and online ads, leveraging tools like Google Analytics to gain insights and optimize future efforts. Compiled membership data, and other data queries using SQL from alumni data lake in Snowflake. Designed and implemented data visualizations to relay essential membership information concisely using both Tableau and Python.

-Designed and implemented data visualizations to relay essential information concisely using both Tableau and Python

Canaan Valley, Wv

For two winters during my undergraduate education I was a part time ski instructor. During this time, I was tasked with instructing all types of clients: children, adults, families, school groups, workshops, etc. I maintained good standing with the PSIA during this time, as well as passed the Level 1 telemark instructors exam.

Morgantown, Wv

-Formulated methods to analyze forensic trace evidence on duct tape samples-Conducted relevant statistical analysis on collected data, helping to prove the method viable-Maintained proper lab notebooks, and managing lab experimentation data was vital to this research- Conducted a presentation of research, as both a poster and seminar, at MAFS conference in Cincinnati (Aug 2017) This research led to a published paper (see attached).

Master Of Science - Ms, Statistics

Relevant Coursework included Machine Learning, Statistical Programming, Theoretical Statistics, Bayesian Statistics

Bachelor Of Science - Bs, Industrial Math And Statistics, Gpa: 3.86

- B.S. in Industrial Mathematics and Statistics with an area of emphasis in Statistics - Graduated Summa Cum Laude - Maintained.