Hello! My name is Angel Lores, and I am a graduate student at Portland State University pursuing a Master's degree in Computer Science.
I enjoy working on Cloud Infrastructure, AI and Machine Learning, and building tools that make complex systems easier to understand and use.
The Standard Insurance Company
Jul 2025 - Present
Building a vector-search based Retrieval Augmented Generation (RAG) system for contract search using LangChain and OpenAI models. Modernizing legacy internal tools into cleaner, more maintainable codebases, often starting from Figma designs and using Copilot to accelerate refactors. Gaining hands-on experience with Azure, AWS, and Google Cloud across development and deployment workflows.
LegitScript
Jan 2025 - Jun 2025
Worked on the merchant monitoring platform, maintaining and extending TypeScript services running on AWS. Safely removed unused legacy code by replacing its behavior where needed and validating that no dependencies were left behind. Added new features and data flows for upserting records into DynamoDB to support internal teams.
The Standard Insurance Company
Jul 2024 - Dec 2024
Built an automated test pipeline so the team no longer had to run long manual regression suites on weekends. Helped maintain Azure infrastructure for internal applications while learning core cloud engineering practices. Contributed to documentation and monitoring improvements for the services I supported.
Fit a smooth cubic B-spline through clicked or file-loaded data points.
C program that allows the user to click points or load them from a file and then fits a smooth cubic B-spline curve through the data. I implemented Gaussian elimination with partial pivoting to solve the linear system for the spline coefficients and used FPToolkit to render the curve and control points.
Tools: C, FPToolkit, Gaussian elimination, cubic B-splines
Neural network from scratch to classify 28x28 pixel digits.
Built a two-layer multilayer perceptron from scratch in Python to classify MNIST digits. I implemented forward propagation, backpropagation with a sigmoid activation, and momentum-based gradient updates, then plotted train/test accuracy and generated a confusion matrix to analyze performance.
Tools: Python, NumPy, Matplotlib, Keras MNIST loader, backpropagation
Recursive fractal drawing with position-based color gradients.
C program that draws an equilateral triangle and recursively subdivides it to generate the Sierpinski triangle using a midpoint algorithm. I wrote a custom line-drawing routine that colors each point based on its position, producing a smooth gradient across the fractal and practicing recursion and basic computer graphics.
Tools: C, FPToolkit, recursion, custom line rasterization
If you'd like to get in touch about opportunities or projects, you can send me a message with this form.