Elijah Grubbs Email & Phone Number

Chicago, Illinois, United States

Ann Arbor, Michigan, United States

Cracking Euchre with Reinforcement Learning.Created a reinforcement learning agent capable of playing euchre. Capable of winning 70% of the time.

Ann Arbor, Michigan, United States

SEC Insider Trading Data AnalysisCreated webscraping tool to collect over 100,000 records of SEC Form 4 filings.Analyzed data to discover trends and behavior driving insider trading behavior

Ann Arbor, Michigan, United States

The goal of this project is to research and replicate state-of-the-art machine learning technologies as described in research papers. I contributed as both the overall project lead and a sub-group lead.My responsibilities as Project Lead included:• Organizing and running weekly project meetings• Keeping up-to-date on all subgroup's progress• Writing the final report detailing the work completed by each subgroup• Creating the slide deck for the MDST Project Presentation showcase My responsibilities as Reinforcement Learning sub-group lead included:• Researching increasingly complex Reinforcement Learning topics throughout the semester• Creating interactive labs coverings topics such as Monte Carlo Control and General Policy iteration• Fostered a deep understanding of Reinforcement Learning in my peers through a combination of guided examples and interactive labs

Ann Arbor, Michigan, United States

The goal of the project was to create an end-to-end reusable pipeline to preprocess data to visualize the stock trades of senators and house members. While also contributing to the work technically, my responsibilities included:• Overseeing project progression and efficiently allotting tasks to speed up deliverable development• Conveying unfamiliar, complicated topics such as stock options, and position basis to group members with non-English first languages, which positively affected independent and group work• Stay up-to-date with members and their interaction with the project, to facilitate a stronger group dynamic and to make sure no one was left behind.

Tampa, Florida, United States

• Developed end-to-end model training pipeline using AWS Glue for ETL processes and Amazon SageMaker for model training, accelerating development lifecycle and ensuring seamless data flow from preprocessing to model deployment.• Streamlined Docker-based deployment pipelines, optimizing containerization leading to a 90% reduction in machine start-up time, enhancing model iteration speed.• Delivered impactful presentations to non-technical stakeholders, effectively conveying project outcomes through visual metrics, fostering cross-departmental collaboration.

Tampa, Florida, United States

• Engineered several features by analyzing over $50 Billion in transaction volume which boosted model performance by more than 400% compared to existing controls in that space.• Presented findings of machine learning models to senior management using quantitative analysis and creative visuals in a combined PowerPoint presentation.• Leveraged several AWS solutions to support real-time model deployment, including Interactive EMR and SageMaker.

Powell

Data Science, Engineering, Third Year

Activities and Societies: Michigan Data Science Team Michigan Poker Club

12

Activities and Societies: Global Scholars Diploma Program, Varsity Swim Team, Junior State of America, Science Olympiad, National Honors.

Education record

Pre-Collegiate Studies, Data Structures, Summer of 2019