AI trade search

Filter builders assume you already know the question. Real review is messier than that.

You might ask, "What went wrong on Tuesdays?" or "Show me the longs I rushed after the open." Those questions need a conversation, not a form.

AI trade search answers with real trade rows, not generated commentary. You can inspect the snapshots, tags, R, notes, and playbook context behind every result.

How it works

Type a question from the trade library or search surface. A large language model maps your wording onto your trades, tags, instruments, timestamps, playbooks, and notes — then returns matching rows with links to snapshots and context, and lets you refine the cohort.

What becomes possible

• Ask questions that would take too long to build as filters
• Move from a hunch to the exact trades behind it
• Inspect real rows instead of trusting a prose summary
• Build review cohorts for analytics, coaching, or playbook cleanup

Where it fits

AI search sits on top of the trade library. Use it when you are exploring your own trades. Use studies when you want to ask a question about a market event across history, and analytics when you already know the scope and want structured reports.

Start here

Ask one question you would normally avoid because the filtering is annoying: "Show me losses after the first hour on ES from the last month." Open the rows before you draw the conclusion.