Joey Hersh

Data Engineer III

USAA / 2024 - Present

• Build and maintain high‑volume data pipelines (~26M records/day) that power credit‑risk and payment‑security applications.
• Delivered a migration of a key risk dataset from a business‑managed tool into the enterprise data platform,improving data quality, lineage, and auditability for regulated reporting.
• Automated model generation: created Python utilities to turn written data specs into staging/load models and validate them against SQL equivalents,reducing manual effort and errors.
• Owned feature work across account‑level risk views and payment‑protection logic; ramped quickly on mid‑stream handoffs and kept delivery on schedule.
• Pitch in on security/maintenance tasks when bandwidth allows to reduce backlog and keep compliance strong.

Founder

Guadalupe Automation / 2025 – Present

• Build AI automation systems for a range of clients, from automated lead generation and customer-support automation to audit-friendly internal operations.
• Install private automation environments (n8n, Make, and agentic AI stacks) for KYC, agentic KYC review, workflow orchestration, and secure human-in-the-loop approvals.
• Support AWS infrastructure maintenance, migrations, and reliability work while training client teams to identify and eliminate busywork permanently.

Co‑Founder & Chief Marketing Officer

Fit & Flourish / 2024 – 2025

• Built an autonomous ad‑creative system using Meta Ads API, n8n, OpenAI image generation, and a multimodal supervisor with a light human‑in‑the‑loop (Discord). Result: 40,000 leads at ~$0.29 CPL (~1/5 the industry average).
• Owned marketing analytics and reporting—pulled performance via API, ran aggressive A/B tests, and auto‑deployed new creatives based on real‑time results.
• Developed the camera‑based sizing calculator: generated a 3D body mesh (Meshcapade) and computed bust/underbust from geodesic paths on the mesh surface to map measurements to size tables.
• Implemented proactive AI customer outreach: MailChimp‑driven surveys and automated follow‑ups to conduct lightweight interviews and increase engagement.
• Owned the e‑commerce site, referral flow, and ad‑tracking dashboards underpinning launch operations.

Chief Executive Officer, Machine Learning Engineer

ReCap Budget / 2021 - 2023

• Raised over $46,000 in seed capital and ~$60,000 in in‑kind prizes (Stumberg Venture Competition).
• Led 7 developers building a serverless, cloud‑native app across 25+ AWS services.
• Built a receipt pipeline (S3→Lambda→Textract) and a fine‑tuned GPT‑3 multiclassifier to assign 100+ categories; structured JSON responses before model function calling was common.
• Implemented security practices aligned to ISO‑style controls for account/transaction data access.
• Shipped LLM‑powered advice features (GPT‑4/LLaMA2) with evaluation loops to keep quality high.

Data Engineering Intern

USAA / Summer 2023

• Built and demoed a secure, on‑prem RAG chatbot (Falcon‑40B + ChromaDB) to 500+ developers; containerized to run fully inside the intranet.
• Created a tool for fast access to underwriting data plus validation scripts,speeding future feature work.
• Worked across Snowflake, dbt, and scheduling tools; helped other teams deploy and adapt my code.

Bachelor of Science in Computer Science

Trinity University / 2020 - 2024

Computer Science Principles I & II, Discrete Structures, Data Abstraction, Functional Languages, Artificial Intelligence, Image Classification, Software Engineering, Computer Design, Computer and Network Security, Principles of Algorithms, Calculus I-III, Differential Equations, Linear Algebra, Probability and Statistics, Physics (Mechanics, E&M), Mechanics (Statics, Dynamics), Mass and Energy Balances, Electric Circuits, Electronics, Engineering Design I-IV, Public Speaking.