Full-Stack Developer

Hi, I'm Matt Maitland

I build web applications, solve technical problems, and make music. From full-stack development to machine learning to creative side projects.

View Projects GitHub

Selected Work

Featured Projects

Professional work and selected side projects, from client applications to interactive experiments.

StringFlux

Most recent active project focused on practical string-processing workflows and cleaner transformation pipelines for day-to-day development tasks.

Problem: Needed a flexible way to test, refine, and reuse string transformation logic without rewriting one-off scripts each time.

Role: Solo developer, concept, architecture, implementation, and iteration.

Outcome: Current in-progress build used as a proving ground for cleaner utility design, repeatable workflows, and stronger developer ergonomics.

TypeScriptUtilitiesDeveloper Tooling

Code

Portfolio Website

Production Next.js portfolio with MDX blogging, a protected admin inbox, and an abuse-resistant contact flow. Built with typed content/data structures and deployment-ready environment management.

Problem: Needed a credible public portfolio that could showcase work, accept contact reliably, and support iterative updates without breaking production.

Role: Solo developer, system design, UI implementation, data modeling, auth, deployment, and documentation.

Outcome: Shipped a live site with server-side validation, honeypot plus Redis rate limiting, GitHub OAuth admin gating, and draft-post protection.

Next.jsTypeScriptTailwind CSSPrismaAuth.jsUpstashMDX

Code

Snake Detector (CNN)

Image-classification pipeline for snake species using a custom CNN workflow: dataset preparation, preprocessing, training/evaluation loops, and inference scripts.

Problem: Needed a repeatable way to classify snake photos under real-world data constraints like inconsistent image quality, class imbalance, and noisy samples.

Role: ML engineer, data curation, preprocessing strategy, model and training setup, and error analysis.

Outcome: Delivered an end-to-end computer vision prototype with reproducible training/evaluation flow and a clear path for future model tuning.

PythonMachine LearningCNNComputer Vision

Code

View All Projects