September 28, 2026 · 8 min read
I spent a month testing Jev, a fast and cheap decision model, as the engine for an agent that watches websites. Here's where it lost, and the approach that ended up working better.
- AI engineering
- Agents
- Research methodology
Read note →August 17, 2026 · 7 min read
The trading ideas that cost me the most were the ones that looked like they worked. Here are the checks I now run on every idea, and two that nearly got through.
- Research methodology
- Quantitative finance
- Statistics
Read note →August 17, 2026 · 4 min read
I moved my portfolio demos from Heroku to Vercel to cut hosting from more than eighty dollars a month to about fifteen, and made sure every demo link still worked afterward.
- Web development
- Deployment
- Operations
Read note →August 17, 2026 · 4 min read
My trading dashboard shows simulated and real-money accounts side by side. Most of the work was making each page clear about which is which and what its numbers measure.
- System design
- Trading tools
- Data engineering
Read note →