Trade smarter with your
AI co-pilot

Smart Trading X is an AI-powered options research tool. It scans option chains on a schedule, filters candidates with rule-based risk gates, and asks AI to research each one. You get a clear proposal with a verdict, a 0–100 confidence score, a thesis and the risks.

Only if you choose to. You pick a mode: paper only, approve each live trade, or auto-run live trades above a confidence level you set. Your capital, trade-count and loss limits apply in every mode.

Yes. Every account has a paper ledger that simulates fills and fees, so you can follow the proposals and see how they play out before connecting a broker.

No. Smart Trading X is a research and decision-support tool. It is not a broker or an investment adviser, and options trading involves substantial risk.

Premium-selling strategies on short-dated options, 0 to 14 days to expiry. The core strategy sells cash-secured puts on liquid stocks and ETFs, with strikes roughly 20–30% below the current price and an absolute delta of 0.15 or less. A second strategy builds put credit spreads, including on the SPX and NDX indexes, with tighter delta caps and a minimum credit-to-width ratio. Leveraged and inverse ETFs are excluded.

Each contract gets a required safety distance, and the strike must clear the largest of three tests: the base out-of-the-money band, a multiple of the options-implied expected move (spot × IV × √(days ÷ 365)), and a technical “safe strike” built from recent volatility (HV20), average true range (ATR14) and nearby support.

Every ticker gets a nightly technical snapshot from daily bars. The gates cover trend (price above a rising 21-day EMA, which sits above a rising 50-day SMA), momentum (RSI in a healthy band, neither oversold nor overbought), structure (price above recent support), Fibonacci retracement, and anchored VWAP from recent cycle highs and lows. Each gate can be switched on or off and tuned per strategy.

Two ways: premium as a percentage of the strike, and return on collateral (premium ÷ the capital set aside). Contracts must clear minimum floors on both, for example a premium of at least 0.4% of the strike. Pricing is conservative: the last trade is used only when it is at or below the ask, otherwise the bid-ask midpoint.

Any contract whose expiry spans an earnings date is skipped, and so is any stock without a confirmed next earnings date. Liquidity checks require a real bid, a reasonable bid-ask spread and quotes no older than 15 minutes. Data that doesn't come from an executable feed is labelled as such.

When a short put moves deep in the money, Smart Trading X looks up to six weeks ahead for a roll: buying back the current put and selling a later, lower strike for a net credit. Suggestions are ranked by return on collateral, priced at the ask to buy back and the bid to sell, so the numbers are never flattering.

Every strategy is a versioned set of tunable parameters: expiries, strike band, delta cap, expected-move multiple, technical gates, earnings rules, return floors, roll rules and AI thresholds. You can clone a strategy, change parameters and dry-run it against today's option chain to see how many contracts each gate rejects. Versions are then compared on their tracked outcomes, so changes are judged on data rather than hunches.

Deliberately conservative. Sell-to-open orders fill at the bid and buybacks at the ask, fees of $0.65 per contract are charged, positions are marked to market every five minutes during market hours, and expiry is settled properly: out-of-the-money puts expire worthless and in-the-money puts are assigned at the strike. Take-profit, stop-loss and time-based exit rules run on the same marks.

Every proposal is also simulated with one contract in a system account, whether or not anyone takes it. That way strategy statistics such as win rate and average return can't be skewed by which trades users happened to choose.

Candidates that pass the rules are reviewed independently by two leading AI models, so their judgments can be compared. Each model receives the technical snapshot, earnings timing, recent news and dealer-positioning context (such as gamma exposure and put walls), and must return structured output: a verdict, a 0–100 confidence score, a thesis, the risks and its sources. Every prompt and response is stored for audit.

A Python quant engine for indicators, options metrics and Greeks; scheduled workers for scanning, nightly technicals, AI review, paper marking and order routing; and a Postgres database. Brokers connect through their official APIs with OAuth, so we never see your password, and tokens are encrypted at rest. Live trading requires two-factor authentication, and every action is written to an audit log.