Nathanael Garza

• Led engineering team of 9 to architect and implement a Kafka service enabling data integration across internal applications using Java Spring Boot, AWS, and Confluent Cloud
• Developed a React UI with TypeScript, Tanstack Query, and OKTA authentication, enabling self-service topic management and producer/consumer configuration
• Spearheaded development of a full-stack React/Spring Boot application that modernizes the management of car dealership configurations which was demonstrated at NADA Show 2024
• Successfully led all aspects of project management including sprint/feature planning, multiple contract deliverables, stakeholder communication, and cross-team collaboration for multiple projects
• Reduced data-integration times by several months by creating a comprehensive application onboarding strategy now adopted company-wide at CDK Global

• Created an admin dashboard for managing canvasing events and volunteers using TypeScript, Next.js, React, Tailwind CSS, and Figma
• Created project roadmaps, architecture diagrams, and API documentation for multiple projects to expedite developer onboarding time
• Conducted pair programming sessions with other developers to improve code quality and team knowledge
• Refactored API endpoints to be more testable as it was previously cumbersome to write tests
• Established new CI/CD pipelines for running E2E Cypress and Jest tests

• Modernized legacy ETL workflows using AWS services (Glue, Step Functions, S3), reducing automotive dealership onboarding from months to weeks and cutting costs by $2M annually
• Enabled non-technical experts to manage dealership databases without engineering support by developing a no-code React UI for creating JSON schemas using JavaScript
• Transformed PICK data into CSV/JSON formats by creating AWS Glue Python scripts
• Automated the updating and testing of our AWS Glue scripts using GitHub Actions and Pytest
• Created Spring Boot BFF microservice for managing AWS ETL workflows using Java AWS SDK
• Identified and remediated SQL injection vulnerabilities in legacy PHP and Perl applications

• Improved Kubernetes cluster management by centralizing AWS EKS info within a backend microservice using TypeScript, PostgreSQL, Node.js, and Swagger
• Automated platform onboarding process reducing wait times from ~24 hours to just a few seconds by disabling repository settings for a short duration using the Stash API
• Introduced a backend service that calculates certain key performance indicators for company executives to evaluate a team’s efficiency using the Bamboo, Jira, and Stash APIs
• Modernized a terminal-based application for setting up CI/CD pipelines to work as a ReactJS web application with a REST API
• Wrote unit and user interface tests using Jest, Mocha, Selenium, and Sauce Labs
• Developed a remote-controlled laser turret in C++ for an outreach video to persuade youth to pursue STEM
• Worked in Agile environment with scrum meetings and 70% code coverage requirements

Gone with the Fire is a wildfire monitoring system built using a distributed LoRaWAN network. The system collects data from various sensors deployed in the field and sends it to an Azure webserver for processing and analysis. The data is then visualized on a web application, allowing users to monitor the status of the sensors in real-time.

The project was developed as part of SCEHacks 2021, where it won several awards, including:

• 1st Place Overall ($500)
• Best Use of Azure

Smart Flask is an IoT water bottle that tracks water consumption via an ESP32, ultrasonic sensors, and an IMU. It sends that data to our deployed React/Node/MySQL services hosted on GCP via CICD pipelines. We have a UI that allows users to set goals and track their water consumption.

The project was developed as part of CruzHacks 2022, where it won several awards, including:

• Best Health Hack
• MLH’s Best Hardware Hack
• QB3’s Best Big Data for the Improvement of Health Care ($1000)

A smart bicycle system that can detect oncoming cars using machine learning. The system uses a camera mounted on the bicycle to capture images of the road. These images are then processed using a machine learning model to detect the presence of cars. If a car is detected, the system sends an alert to the cyclist, allowing them to take appropriate action to avoid a potential collision.

The project was developed as part of HackDavis 2022, where it won the following award:

• Most Technically Challenging Hack

A mini monster truck that can collect air and soil data. The data is then sent to an AWS webserver for processing and analysis. The system is designed to help farmers monitor the health of their crops and make informed decisions about irrigation and fertilization.

The project was developed as part of MakeUC 2022, where it won the following award:

• Best Use of Web Automation Technology

An IoT barbell collar that automatically records number of reps and time under tension. The device utilizes a Raspberry Pi Pico W, an IMU, and a custom 3-button OLED interface that allows users to select exercises and weight. It automatically tracks workout metrics and integrates with a full-stack TypeScript/React/Node/MySQL application deployed on AWS infrastructure via CICD pipeline. The web application provides a user-friendly interface for users to view their workout data, track progress, and set fitness goals. The system is designed to help users optimize their workouts and improve their overall fitness.

The project was developed as part of CruzHacks 2023, where it won the following award:

• Best Health Hack

A full web browser based serial telemetry tool for reading and writing data to microcontrollers. The tool was built to be responsive using only vanilla CSS. It is used in SJSU Robotics and San Jose State University embedded system courses. The tool is open source and available on GitHub.

1. Test Driven Development by Kent Beck
2. Refactoring by Martin Fowler
3. Clean Coder by Robert Martin