Work Experience

Engineer I

Jan 2024 – Present

XYPRO Technology

  • Designed and developed a full-stack RAG system serving 50 internal users across development, support, and management — ingesting ~18,000 documents (product manuals, guides, customer case summaries, Confluence pages, KB articles) from Salesforce, Jira, Confluence, and Slack via event-driven triggers
  • Built FastAPI backend integrated with Gemini LLMs via LangChain and vector store retrieval; deployed with Docker Swarm + Traefik on GCP with MongoDB
  • Developed secure Vue3 + TypeScript frontend with role-based UI and Entra ID authentication; added MCP layer enabling AI agents to connect via API keys with usage tracking
  • Led XYGATE SecurityOne UI migration from Vue2 to Vue3
  • Leading initiative to migrate test suite (ReadyAPI, Rest-Assured, Selenium, WebDriverIO) to unified Playwright + TypeScript suite, leveraging AI coding agents and training the team on the process

ML Research Assistant

Oct 2023 – May 2024

California State University, Northridge

  • Trained DaViT, EfficientNetV2, and MobileOne models for multi-class pomegranate disease classification using transfer learning; achieved 99%+ accuracy across five classes
  • Co-authored IEEE-published research paper presented at the 2025 IEEE International Conference on Computing and Machine Intelligence (ICMI)

Cloud Support Associate Intern

Jun 2023 – Aug 2023

Amazon Web Services (AWS)

  • Completed 12-week training program on AWS cloud services (EC2, S3, load balancers); worked through customer case simulations with seasoned engineers
  • Earned AWS Solutions Architect – Associate certification during tenure

Program Manager & Manager

2013 – 2023

Computerwisekids Inc.

  • Managed curriculum across 36 sites; led full transition from in-person to online instruction for 500+ teachers; grew from Manager to Program Manager overseeing social media, staff development, and onboarding systems

Publications

IEEE 2025

Pomegranate Disease Classification Using Advanced Deep Learning Models

J. Sanchez, J. Enriquez, A. Verma

2025 IEEE 4th International Conference on Computing and Machine Intelligence (ICMI), MI, USA · pp. 1–8

Trained DaViT-Base, EfficientNetV2M, and MobileOne-S4 on annotated pomegranate images across 4 disease + 1 healthy category. All models exceeded 99% accuracy (DaViT: 99.28%, EfficientNetV2M: 99.54%, MobileOne-S4: 99.15%). Demonstrates that lightweight models are viable for real-world agricultural disease detection on mobile devices and drones.

99%+ accuracy Lightweight models for mobile & drone deployment
DOI: 10.1109/ICMI65310.2025.11141088

Skills

AI / ML

PyTorch LangChain RAG Transfer Learning Vector Stores Gemini LLMs AI Coding Agents

Backend

FastAPI Python Docker MongoDB GCP

Frontend

Vue3 TypeScript React Native JavaScript

Cloud & Infra

AWS (SAA-C03) GCP Docker Swarm Traefik

Testing

Playwright ReadyAPI Selenium

Other

Git SQL C C++ Java

Education & Certifications

B.S. Computer Science

California State University, Northridge

Graduated May 2024 · Magna Cum Laude

AWS Solutions Architect – Associate

Issued August 2023

Projects

Pomegranate Disease Classification

Multi-class plant disease classifier using transfer learning on annotated pomegranate images. All three models exceeded 99% accuracy. Published in IEEE ICMI 2025.

Python PyTorch Transfer Learning DaViT EfficientNetV2
IEEE Paper

Wohpe App

Cross-platform mobile application developed in collaboration with NASA JPL and CSUN ARCS. Features a PostgreSQL backend and Docker deployment.

React Native PostgreSQL Docker Expo

Pneumonia Binary Classifier

Deep learning model for binary classification of pneumonia from chest X-rays using EfficientNet with transfer learning on the Kaggle chest X-ray dataset.

Python PyTorch EfficientNet Transfer Learning

Real-time Face Filter

Real-time face filter application using MediaPipe for landmark detection, InsightFace for recognition, and OpenCV for rendering augmented overlays.

Python MediaPipe OpenCV InsightFace
GitHub

Online JavaScript IDE

Browser-based JavaScript IDE with live code execution, syntax highlighting via CodeMirror, and a clean React interface for learning and prototyping.

React CodeMirror JavaScript HTML CSS
GitHub