I am Linus, building private AI systems where stocks, live market context, model routing, and execution workflow meet.
I am not interested in generic AI wrappers or polished demo software with no operating depth behind it. My focus is building tools that sharpen real market decisions before, during, and after execution.
- swing-to-day trading focus
- signal quality over noise
- market context over isolated prompts
- private systems before public packaging
- execution discipline over automation theater
- Linux, Debian, Git, GitHub, VS Code
- builder-first mindset with fast iteration
- AI-native workflow with live model testing
- earlier FiveM scripting background with shipped work
- systems built for actual use, not presentation
A private finance AI workspace built around the way I actually trade: charts, catalysts, headlines, earnings, memory, risk, and execution context in one operating surface.
The project is a live private platform, not just a finance-themed assistant concept. It combines a custom trading desk UI, persistent chat, finance-focused reasoning modes, memory, and a workflow built around actual market use instead of generic chat.
- reduce noise before execution
- keep finance context attached to the decision
- track chats and evolving reasoning over time
- build a sharper link between research, memory, and trades
- create a private operator stack before deciding what should ever be public
Python JavaScript HTML CSS Flask HTMX PostgreSQL Chroma OpenRouter Gemini
The stack has moved forward from the older setup. The current system runs on Gemini 3 Flash through OpenRouter, with different product modes tuned around finance workflow instead of one flat assistant behavior.
