Why Claude?
Claude isn’t a tool I use occasionally. It’s integrated into how I work every single day. Research, learning, development, architecture decisions, content writing, code review, debugging, and exploring ideas. It’s the closest thing to having a senior colleague available 24/7 who can context-switch between Laravel DDD patterns, NixOS configuration, PostgreSQL query optimization, and infrastructure planning without missing a beat.
How It Fits Into Everything
Claude runs across multiple interfaces depending on the context — Claude Code in tmux for hands-on development, the web interface for research and longer conversations, and the API when building AI-powered features for clients.
Where it shows up
- Development: Claude Code as a pair programmer inside Neovim + tmux sessions, writing code, reviewing diffs, and catching issues
- Research: exploring new technologies, understanding documentation, evaluating architectural approaches
- Learning: deepening knowledge in areas like Go, NixOS, and networking concepts
- Content: writing, structuring, and optimizing content like this very site
- Architecture: talking through DDD patterns, database schemas, API design, and system boundaries
- Client work: building AI-integrated features using the API for projects like Dadooo.ai
Where it stops
Sensitive client environments with strict data policies stay offline. When the work involves proprietary data that can’t leave the client’s infrastructure, Claude doesn’t see it — that’s a boundary I maintain deliberately.
Status
Active: daily driver for research, development, learning, and creative work.