What "AI day trading" means here

AI Day Trading Without the Always-On Bot

Cortiq's agent wakes on a trigger, runs a full research-to-execution cycle, and trades one instrument. Its guardrails live outside the model, not inside a prompt.

The frequency model

What AI day trading usually means, and what it means here

"AI day trading" conjures a bot firing nonstop. That is not the model Cortiq runs.

Isn’t a scalping bot spraying orders every tick.

Is event-driven: it acts only when a trigger you defined fires.

Isn’t scanning dozens of symbols for setups.

Is one instrument, all day, in depth.

Isn’t "more trades equals more edge."

Is as few trades as the playbook justifies.

Isn’t latency-driven or high-frequency.

Is candle-timescale decisions a human can audit.

What wakes the agent

Three trigger types bound how often it can act

Candle close

Wakes when the timeframe candle it watches closes, not before and not on every tick.

Price level

Wakes when price reaches a level you or the playbook defined, then goes back to sleep.

Scheduled time

Wakes at a time you set: a session open, a news window, a fixed check-in.

Between triggers, the agent sleeps. It is not polling the market, and that gap between wakes is what keeps this day trading instead of constant firing.

See a full cycle, step by step →
110-second overview

See the whole loop, start to finish

A narrated tour of the one-market loop — research the edge, write the playbook, backtest it on real history, then go live — with every AI call and every guardrail on screen.

cortiq · session walkthrough01:50
Chapters
Narrated overview · 1:50
Produced from the desktop app on a single instrument. Jump to any chapter.
produced · Jul 2026· v1.x· Windows desktop
Frequency guardrails

What caps how often it can trade

Approval gates

Mutating actions wait for you unless a session is explicitly set to run unattended.

Restricted hours

Confine the agent to specific trading windows; outside them it won't act.

Risk limits

Enforced by a validator layer outside the model: per-symbol caps, drawdown limits, and a loss-streak stop built to catch overtrading on a bad day.

None of these three levers live inside the model's prompt. The model cannot argue its way past them.

cortiq · risk · guardrails
Cortiq risk guardrails panel showing per-symbol caps, drawdown limits and a loss-streak stop enforced outside the AI model
Risks of frequency

Be conservative about how often it trades

Trading financial markets carries a high degree of risk and may not be suitable for all investors. You can lose some or all of your invested capital. Read the risk disclosure.

Watch outEvery trade pays spread and commission; higher frequency compounds cost, and a real edge has to clear it. Cortiq doesn’t lower your broker’s costs.
Watch outEach wake spends an AI API call. More frequent triggers mean more provider cost you pay, separate from Cortiq (BYO-AI).
NoteMore trades doesn’t mean more edge. Overtrading can turn a positive expectancy into a negative one; trigger and playbook design are what prevent it.
NoteRun a paper session first. Watch how the agent’s cadence behaves before trusting it with an unattended live one.

Trading financial markets carries a high degree of risk and may not be suitable for all investors. You can lose some or all of your invested capital. Read the risk disclosure.

See the full mechanics

The full walkthrough behind this page: the complete trade cycle, the validator layer, and what Cortiq deliberately does not do.

See the full mechanics
Ready to run it yourself

One-time activation key, bring your own AI, 14-day refund, free v1.x updates.

Buy an activation keyOne-time · 14-day refund · 2-machine activation
One market · one disciplined agent

Put a disciplined agent on one market.

Buy once, own forever. Bring your own AI, keep the risk logic outside the model, and watch every decision before it touches your account.

Read the install guide
One-time · 14-day refund · 2-machine activation · free v1.x updates