Joey Hersh
Trinity University Computer Science Student
Email: jhersh@trinity.edu
Phone: 713-505-3289
Website: resume.joeyhersh.com
GitHub: joey12725
Programming Skills
- Python (+Pandas, PyTorch, TensorFlow…)
- Scala
- C++
- Haskell
Cloud Skills
- SageMaker
- Managed ML (Textract, Comprehend, Bedrock, etc.)
- S3, Athena
- DynamoDB
- EC2, RDS, VPC
- Lambda
- SQS, SNS
- IAM, Cloud Watch
- Amplify, Cognito
Other Skills
- Mathematical Modeling in MATLAB, R studio, and Python
- CAD in Fusion360
- Full Adobe suite
- Project/team management
Work Experience
USAA (Summer 2023)
Data Engineering Intern
- Collaborated on an agile team to manage end-to-end development cycles, driving a new feature from requirement inception through to production, utilizing tools such as Snowflake, DBT, Talon, and Control M.
- Created a a specialized tool to enable the business to rapidly access and analyze credit card underwriting data, and developed data validation and automated testing scripts to enable faster development of similar features in the future
- Presented to an audience of 400 developers and data scientists from across the organization in a speaker series called Let’s Talk Data – Generative AI where I demoed a vector database backed chat bot powered by Falcon-40b I built to improve the experience of using the USAA intranet to find obscure information
ReCap Budget (2021 - 2023)
Chief Executive Officer, Machine Learning Engineer
- Raised $46,000 in seed capital and cash prizes, and over $60,000 of in-kind prizes through the Stumberg Venture Competition.
- Lead a team of seven developers building and implementing full architecture for a (mostly) serverless, cloud native application that utilizes over 25 AWS services.
- Managed communication between the development and business units, explaining complex backend logic to people with minimal expertise.
- Met cyber-security standards for Plaid production-level access (iso 27001) to account balance and transaction data from most major American banks
- Leveraged ensemble learning algorithms (XG-BOOST) as well as foundational LLM models (GPT-4, LLaMA2, finetuned GPT-3.5 Turbo) alongside AWS-managed ML services such as Textract, Comprehend, etc. to analyze user uploaded receipts and generate unique advice on how to spend better.
Breakthrough CTX (Summers of 2017-2019)
IT Support and Systems Administration for Staff and Students
- Summer volunteer position through St. Andrew's Episcopal School assisting Breakthrough, a non-profit that helps sixth through twelfth-grade students that hope to become first-generation college students achieve their goals.
- Provided IT support each summer to both staff and students including mobile device repair, staff software deployment, managing student system permissions, campus network management, AV setup and repair, etc.