Projects

What it does A Python GUI app for sorting files through a Tinder-style swipe interface. Swipe right to keep, left to delete — making file cleanup faster and more intuitive than traditional file browsers. Why I built it File organization is tedious. Clicking through folders, selecting files one by one, confirming deletions — it’s slow, error-prone, and honestly just lame. Almost nobody wants to clean up their files. I wanted something different: show me a file, I decide in a second, next. Make it almost fun, or at least less painful. ...

What it does An AI character companion with personality, memory, and journaling capabilities. Text-first design, voice optional. Self-hosted and privacy-focused. Why I built it Since LLMs first emerged, I’ve been fascinated by them. We suddenly got a huge step closer to the AI assistants we see in media — think J.A.R.V.I.S. I knew I had to build my own version. When tool use became available, I got obsessed with the idea of controlling everything through natural language. Imagine the accessibility implications: an LLM that could control Home Assistant, manage your files, automate your workflows. ...

What it is My home infrastructure: a Proxmox virtualization host behind an enterprise-grade firewall, running everything from home automation to media services. Architecture Internet (105/39 Mbps VDSL) │ ▼ ┌─────────────────────────┐ │ OPNsense Firewall │ │ (Sophos SG 210) │ └─────────────────────────┘ │ ├── LAN (192.168.128.0/24) │ └── Daily devices │ └── LAB (192.168.129.0/24) └── Proxmox Host ├── Home Assistant ├── Media services └── VMs & containers Key components Network & Security Firewall: OPNsense on Sophos SG 210 hardware DNS: DNS-over-TLS with Quad9 (primary) and Cloudflare (backup) Segmentation: VLANs separating daily-use from lab environment Monitoring: Gateway monitoring, custom LCD status display on the firewall Virtualization Hypervisor: Proxmox VE Containers: LXC for lightweight services VMs: For workloads that need full isolation Home Automation Platform: Home Assistant Protocol: Zigbee (thermostats, sensors) Future: Integration with eva (my AI assistant) Design principles Self-hosted first — Avoid cloud dependencies where possible Security by default — VLAN isolation, encrypted DNS, MFA everywhere Document everything — If I can’t reproduce it, I didn’t learn it Learning platform — Real infrastructure to experiment with What I learned Enterprise firewall configuration and hardening VLAN design and inter-VLAN routing DNS privacy (DNS-over-TLS, DNSSEC) Proxmox storage and container management Home automation protocols (Zigbee, MQTT) Running servers costs more energy than you’d think Roadmap IDS/IPS with Suricata WireGuard VPN for remote access Monitoring dashboard with Grafana Custom ESP32-H2 Zigbee devices

What it does A Discord music bot for playing music from YouTube and Spotify1 in voice channels. Full player controls — play, pause, skip, queue management — built for my friends and me. Why I built it My friends and I wanted seamless music playback during our D&D sessions in Discord. The existing bots either lacked features we wanted or locked them behind premium tiers. I saw it as a chance to learn Java, work with real-time audio streaming, and build something we’d actually use regularly. ...