Badminton Stroke Classifier
Classifies badminton strokes from match video using PyTorch models and a React and FastAPI application. I contributed major frontend, backend and integration work, plus deployment and CI configuration.
Public repositoryCurtis Martin
Software engineer and Care Culture co-founder in Sydney.
I build and maintain the internal systems used across Care Culture and have completed a Master of Information Technology at UNE. I’m seeking a full-time graduate or junior software engineering role.
Engineering experience
Care Culture internal systems
After working in disability support, I co-founded Care Culture and became the sole developer of its internal tools, automations and data workflows.
- Built onboarding, booking, invoicing, reporting and compliance systems using JavaScript, Google Apps Script and APIs.
- Automated onboarding documents, signatures, forms and record creation, cutting hands-on processing to under two minutes.
- Built Booking Buddy and have maintained it since 2021 to manage participant bookings.
Selected work
Projects
svelte-a11y-panel
I rebuilt Care Culture’s website in SvelteKit and TypeScript, then extracted its accessibility panel into a reusable Svelte 5 package.
Project siteInvoice-to-CSV
Converts invoice data from Google Sheets into Xero-ready CSVs for bulk import. It has handled Care Culture’s invoicing since 2021.
Public repositoryBadminton Computer Vision Auto-Annotator
Co-developed a resumable Python pipeline and delivered a versioned research dataset covering 6,833 rallies across 86 match videos, using existing human contact annotations. I built the dataset-builder and export layers, co-designed the schema and benchmarked features. I also aligned commentary and manually reviewed 134 passages to check timing and rally associations.
Team repositoryEducation
Master of Information Technology
GPA 6.94/7 · WAM 90.85
Relevant study: software development, databases, networks and security, distributed computing, artificial intelligence and machine-learning capstones.