// portfolio
Sawyer Czupka
Software Engineer.
Full-stack · ML Systems · Infrastructure.
I build production software across backend engineering, ML systems, and
cloud infrastructure. I'm a backend-first full-stack engineer — API
design, data pipelines, and system architecture are where I'm most at
home, though I'm fully capable on the frontend too.
My ML work spans both applied systems and model development. On the applied
side I've built RAG pipelines, LLM evaluation frameworks, and production model-serving
infrastructure. On the research side I've trained models from scratch — dense
neural networks, graph neural networks, transformers, reinforcement learning
agents, and computer vision classifiers — through my work at William & Mary.
I'm also comfortable managing the full infrastructure stack: Kubernetes clusters,
CI/CD automation, cloud deployments on Azure and Cloudflare, Terraform-provisioned
environments, and self-hosted homelab services.
Recently finished my BS in Computer Science & Data Science at William & Mary,
Cum Laude.
// selected projects
all projects →LapEvo
2025 →iRacing telemetry tracker and AI racing coach — real-time 60Hz data pipeline, React dashboard, and TimescaleDB time-series storage.
Personal Homelab Infrastructure
2023 →Production-grade homelab: 116-pod Kubernetes cluster on Proxmox, GitOps via ArgoCD, Terraform/Packer infra-as-code, and automated TLS.