Current Projects

Home Lab

Started ~2002 Status Ongoing

A fully domain-joined home infrastructure running under the DTPC.US Active Directory domain, spanning multiple physical hosts and services. What began as a single spare machine has grown continuously over more than two decades into a fully managed private network.

Networking was originally handled by a custom pfSense router build before migrating to a full Ubiquiti UniFi stack for improved manageability and performance. Every device in the home is joined to the DTPC.US domain, providing centralized authentication and management across the entire environment.

  • Dell PowerEdge T330 — Primary Hyper-V host running the Primary and Secondary Domain Controllers, internal web hosting, Grafana monitoring, and miscellaneous VMs
  • Custom MicroTower — Secondary Hyper-V host running the Tertiary Domain Controller and game servers
  • Synology DS3617xs — 12-bay NAS with 12× 10TB drives in RAID 6, used for bulk storage, media archiving, and automatic phone backup for photos and videos
  • Intel NUC — Dedicated Plex media server
  • Full Ubiquiti UniFi network stack

sFTPGUI v2

Started ~2014 Status Active Tech C#, WinForms, WinSCP, Renci SSH.NET

A C# WinForms application for syncing remote VPS downloads to local NAS storage for backup and archiving. What started as a simple FTP sync utility has evolved through two major versions — first adopting SFTP via Renci SSH.NET, then migrating to WinSCP for improved reliability and features.

  • Real-time per-file transfer progress including speed and completion percentage
  • Multithreaded transfers for parallel file processing
  • Runs continuously, automatically re-queuing the sync every hour
  • Tracks previously completed transfers to avoid redundant re-downloading

Local AI Lab

Started ~2024 Status Stable / Exploratory Hardware NVIDIA RTX 4080

A locally hosted AI environment running entirely on a personal development machine running Windows 11, powered by an NVIDIA RTX 4080. Set up and explored both large language models and image generation without relying on cloud services.

  • Ollama with Llama for local LLM inference
  • Automatic1111 / ComfyUI with Stable Diffusion for local image generation

Primary motivation was hands-on learning and exploration of self-hosted AI tooling as the space rapidly evolves. All inference runs locally on the RTX 4080 with no external API dependencies.

This Site

Started 2025 Tech Astro, TypeScript, CSS Status Active

Rebuilt from a legacy PHP single-page application to a static Astro site with a modern dark theme, responsive layout, and clean URL structure.