Aparoop
Tickerly · Personal project · Phase 1

I built myself an AI Financial Analyst.

Tickerly is a personal AI project. I trade options actively against my own capital, and the standard toolkit kept letting me down. So I built one that doesn’t. Phase 1 is single-operator validation. What comes after gets decided once Phase 1 teaches me something.

01 Why I built it

I trade options actively against my own capital. The standard toolkit — charting platforms, broker apps, signal services — wasn’t enough.

None of it remembered what worked last quarter. None of it adjusted for the patterns I kept losing to. None of it told me when to stand aside. So I built one that does.

Tickerly is a workbench. The discipline is the product.

It does for me what an analyst would do for an institutional book — but at a scale of one. After enough iteration, the workbench became a methodology: synthesis over signals, memory over hot takes, conviction-graded views. Whether that methodology scales beyond one operator is a Phase 2 question.

02 What it feels like

Sample output from a typical day on the workbench. Some mornings it’s a one-page note. Some afternoons it’s a position alert that arrives before I noticed the move. Some nights it’s a regime warning that lets me adjust before the open. Names and figures below are illustrative, not recommendations.

Morning note Mon · 6:15 AM PT
Three things worth your attention this morning.
→ A name on your watch list reports tomorrow AH. Options are pricing a ±5% move; the eight-quarter average is ±2.3%. Premium is rich. If you’re long short-dated calls, this is the window to take a third off.
→ A laggard in an out-of-favor sector is breaking above its 50-day on rising volume. Leadership is changing beneath the surface. Worth a closer look before consensus catches up.
→ One of your held positions has cleared the technical level you flagged six weeks ago. Original thesis check: still intact?
Position alert Fri · 1:30 PM PT
One of your held positions needs a decision.
→ Binary catalyst is five sessions out. Implied vol is rich; even a positive print could cost you 40-60% of remaining premium on IV crush alone.
→ Recommended: trim roughly a third before the event. Lock part of the realized gain. Hold the rest with a defined stop and a take-profit target.
→ Path-dependent: if the print misses, exit the remainder next morning. If it beats and continues, plan a second trim at the next round level.
Regime shift Thu · 9:00 PM PT
The macro tape changed overnight.
→ Bonds sold off. Dollar firmed. Crypto under pressure. The composite is signaling risk-off — not crisis, but a meaningful regime change you need to register before the open.
→ Three of your four held positions are on the wrong side of the shift. Tomorrow open: trim, hedge, or reduce gross. Don’t add into the move.
→ Two themes from this week are still working independently of the broad de-risking. Worth flagging for re-entry once volatility normalizes.

The output is reasoned, not signaled. You can ignore it. You can argue with it. You can’t catch it not noticing.

03 Three principles
Principle 01
Synthesis over signals.
A good analyst doesn’t just say buy. They explain why, what could break the thesis, and what to watch for. Tickerly produces reasoned views — not green and red arrows. The work is in the synthesis: connecting flow, fundamentals, macro, and your specific positions into one coherent take.
Principle 02
Memory over hot takes.
Every call is logged. Every outcome is measured. Every Sunday at 6 PM ET, the analyst grades itself against what actually happened the previous week. Patterns that drifted off get adjusted. Tiers that over-claimed get demoted. The analyst that ships next week is not the one that shipped this week. That’s the loop. Without it, nothing else matters.
Principle 03
Conviction over confidence.
Confidence is loud. Conviction is calibrated. Tickerly attaches a conviction score to every recommendation — based on signal alignment, prior accuracy on the same pattern, and current data trust. Low conviction means smaller size, tighter stop, or no trade at all. The bravest analyst output is “I don’t know enough yet.”
04 Current state
20
closed trades
80%
overall win rate
+38%
P&L, 35 days

The track record so far on options calls. CALL side is calibrating cleanly; PUT side is underperforming and the system has trimmed its own exposure there until the pattern resolves. Numbers refresh every Sunday 6 PM ET. This is the workbench’s own grade card from running against my capital — not a beta service, not a paper-trade simulation.

05 Where it goes next

Phase 1 is single-operator. Running. Working. I’m learning what holds up under real-money conditions and what doesn’t.

Whether there’s a Phase 2 — a private beta, a hosted version, anything commercial — depends on what Phase 1 teaches me. I don’t have a launch date. I don’t have a waitlist. If those things make sense later, they’ll happen later.

If you want updates on what I learn from the build, follow The Loop. That’s where the patterns get written down as I see them.

Read The Loop → My fractional work ← Home