Joe Brijs

Software Engineer and CS Student

Software Engineer | Student at Utah State University

Resume download

Projects

Neural Network Stock Predictor Web App

Overview

This project involved developing a system to predict stock prices using deep learning and cloud services. It collects historical stock price data, engineers key financial indicators, and stores the processed data in Amazon S3 for easy access. I trained LSTM neural networks using PyTorch to forecast next-day stock prices for leading tech companies, providing timely insights. The pipeline includes AWS Lambda functions that automatically fetch new data and trigger predictions after market hours, managed by AWS EventBridge Cron jobs for consistent updates. Predictions are accessible via API endpoints hosted on AWS Lambda and managed through API Gateway, enabling real-time comparisons with actual prices. The frontend is a dynamic React web app displaying stock predictions and performance metrics, hosted on AWS S3, distributed via CloudFront, and utilizing Route 53 for DNS management.

Technologies:

www.faangfinance.com exit_to_app GitHub Repo exit_to_app

AI Workout Plan Web App

Overview

This web application enables users to input personal information to receive customized workout plans. Users can track and store their workout progress directly within the app. The project utilizes React.js for the frontend, is styled with TailwindCSS, uses Django for the backend, and integrates the OpenAI API to dynamically generate personalized workout recommendations.

Technologies:

GitHub Repo exit_to_app

Habit Tracker Android App

Overview

The Habit Tracker Android App, built with Kotlin and Jetpack Compose, enables users to track 'do' and 'don't' habits seamlessly. Featuring intuitive gestures, smooth transitions, and engaging animations, this app offers a robust mobile experience. Habits are saved persistently across sessions through Firebase integration. Additionally, the implementation of Google Ads enhances the app's realism and functionality.

Technologies:

GitHub Repo exit_to_app

Typing Tutor Game

Overview

The Typing Tutor Game is an interactive, front-end-only typing game built using React.js with TypeScript. The game enhances user engagement through reactive colors and animations that indicate which key to press and confirm the key pressed, helping users to memorize key positions. Additionally, a cursor and prompt animation visually track and highlight the user's progress throughout the game. This game is intuitive and visually appealing.

Technologies:

GitHub Repo exit_to_app

Professional Experience

USU ASPIRE Lab

Student Software Developer and Research Assistant

Nov 2024 - Present

Working on electric vehicle infrastructure optimization through machine learning and full-stack development at Utah State University's NSF-funded ASPIRE research center.

  • Reinforcement Learning Research: Conducting research and developing a Reinforcement Learning framework to optimize electric bus charging schedules, aiming to enhance efficiency and reduce energy costs across Utah's EV infrastructure.
  • ML Model Integration: Integrating Python-based machine learning models into a Rust web API, enabling real-time predictions and ensuring seamless interaction between data pipelines and backend services.
  • Web Application Development: Designing and maintaining new features for an EV reservation web app using Express and React.js, improving user experience and system reliability.

Elektrik -

Early Stage Start Up, B2B SaaS and Electrical Distributor

Software Engineer - Intern

May 2024 - Nov 2024

I worked with Rust in the backend services and React for frontend development. Some of my key contributions included:

  • Search Tools for Medium Voltage Accessories: Designed and implemented robust search tools by analyzing datasheets and integrating part data into the web app. This contributed to a 25% increase in revenue year-over-year by improving part discoverability for customers.
  • Credit Application Process Optimization: Reduced the credit application time from days to minutes by creating an adaptive form tailored to specific business types and payment terms. Automated submission workflows using Slack integration via GCP Pub/Sub, generated dynamic PDFs, and incorporated the entire approval process into the web app dashboard.
  • Angular to React Migration: Led the successful transition from Angular to React, reducing the frontend codebase by 40%, enhancing UI/UX, and improving developer efficiency, which resulted in a more maintainable and scalable codebase.
  • Rust CLI for Data Standardization: Developed a Rust-based command-line tool to standardize MongoDB collections, saving an estimated 20 hours of employee time and improving data consistency.
  • Startup Environment Adaptability: Thrived in a fast-paced, early-stage startup, where I adapted to evolving requirements, contributed to cross-functional roles, and made key product and technical decisions that shaped the company's core offerings.

In-N-Out

Associate & Clean-up Crew

Jun 2022 - May 2024

Provided excellent customer service to the greater Cache Valley area. Was part of the 5am Clean-up Crew, cleaned and prepared the store for opening before heading to 10:30am classes.

Lets get in touch:

joebrijs@gmail.com
Resume download