AI Trading Systems Lab Journal is an electronic journal based on a fully autonomous AI stock trading system run with proprietary capital only. We do not manage client assets.

The journal publishes experience-based operating notes from that reality: what worked, what failed, and what we changed to keep the system stable in live markets.

The focus is not theory. It is the practical work that begins after backtests are complete – live deployment, monitoring beyond performance, change control, incident handling, and the uncomfortable edge where paper trading stops being informative and autonomy becomes difficult to maintain.

We publish because the hardest work in this field is rarely documented: keeping an automated system stable, observable, and auditable under real conditions. What we share is operating discipline: how systems are designed, monitored, controlled, and governed once real capital, real constraints, and real failure modes are involved.

All material is intentionally non-actionable. We do not disclose strategies, traded instruments, features or signals, data sources, or performance forecasts. We publish operating lessons, not tradeable information.

Subscribers receive access to the online archive and to new issues as they are released.