Software I use, gadgets I love, and other things I recommend.

The tools and setup I use to build software, stay productive, and get things done.

Workstation

MacBook Air M2, 16GB RAM
This little machine is my Swiss Army knife of software development. Great ergonomics from anywhere: coffee shops, airports, couches. I've done my most productive work from this laptop. The M2 chip handles everything I throw at it without breaking a sweat.
Main Station - Custom PC
Homebuilt PC with RTX 5090, Ryzen 9950X3D. For the heavy lifting: AI model training, running local LLMs, and anything that needs serious compute. Connected to an Alienware AW3225QF monitor.
KEF LSX II
My audio setup for the main station. Clean, powerful sound that makes long coding sessions more enjoyable. What I really like about it is that it's fully controllable remotely via Cast, WiFi, or Bluetooth, so it doubles up as room audio.
Logitech POP Icon Keys Wireless
I've always been a fan of mechanical keyboards, but lately I've been gravitating toward a more minimalist setup. The POP Icon Keys strikes a nice balance, tactile enough to be satisfying, simple enough to stay out of the way. I also have a Keychron K2 Max.

Neovim
My primary editor. Once you get comfortable with modal editing and build your config just right, there's no going back. Fast, efficient, and endlessly customizable.
Ghostty
A modern, GPU-accelerated terminal emulator. Fast and clean, exactly what a terminal should be.
VS Code
For when I need a full IDE experience or when collaborating with others. The extension ecosystem is unmatched, and sometimes you just need those extra features.
And a ton of other software
Depending on the use case: database tools, API clients, profilers, debuggers.

Obsidian
My second brain. Everything goes in here: notes, ideas, research, project planning. The ability to link thoughts together and see connections emerge over time is invaluable.
Homebuilt AI Assistant
Built using local LLMs and RAG (Retrieval-Augmented Generation). Privacy-first AI assistance that runs entirely on my hardware. I use it for research, code generation, and working through complex problems without sending data to third-party services.