The page we'd want to read before subscribing

01Why this page exists

Most trading products oversell, hide their assumptions, and disappear when their numbers stop working. We've been burned by enough of them. This page is the antidote — written in plain English, not legal copy.

If you decide Falcon AI isn't right for you after reading this, that's a fair outcome. We'd rather you make an informed decision than convert on a hero stat and refund a week later disappointed.

02Who's behind Falcon AI

Falcon AI is operated by Falcon AI LLC, a Florida limited liability company. Behind the LLC is the two of us — I (Marnie) trade Falcon AI live on Topstep every session; my husband and I engineered the strategy together over years of building, testing, and breaking it. We built this for our own accounts first and opened it to subscribers second.

We are not a venture-backed company. We are not a 100-person team with a marketing department. There is no celebrity guru behind this — we sell signals, not a course or a coaching program.

• Entity: Falcon AI LLC, registered in Florida, USA
• Billing: All payments processed by Stripe. We never see or store your card details.
• Support: support@thefalconai.com — replied to by a real person, typically within hours during US trading days
• Live trading: Yes, we trade Falcon AI on our own accounts. We are subject to the same drawdowns and signal quality as every subscriber.

03How the backtest numbers were generated

The headline numbers on the homepage and pricing page are hypothetical backtested results — not live trading records. Here is exactly how they were generated, so you can reproduce them yourself if you have a TradingView Pro account:

The core configs (TITAN · PRIME / TITAN · EXTENDED / TITAN · GOLDEN / RAPTOR · PRIME / RAPTOR · EXTENDED / RAPTOR · GOLDEN) + BTC Elite

Falcon AI is one strategy with core configurations that differ only by timeframe and session window — plus an Elite-exclusive BTC 24/7 config (signals-only). Same core engine, same risk math across all six. We run all core configs live in our own Topstep accounts. Subscribers can run any of them; RAPTOR · GOLDEN has the longest verified track record.

• 🛡️ TITAN · PRIME (30M · 9–12 ET) — prop firm pick · 70.8% WR · PF 2.67 · +$200,489 net · 1.73% max DD · 30 of 30 profitable months (backtest)
• ⚖️ TITAN · EXTENDED (30M · 9–1 ET) — recommended default · 70.2% WR · PF 2.46 · +$228,540 net · 1.71% max DD · 30 of 30 profitable months (backtest)
• 🎯 TITAN · GOLDEN (30M · 10–1 ET) — 70.6% WR · PF 2.06 · +$124,583 net · 1.79% max DD · 29 of 30 profitable months (backtest)
• 🚀 RAPTOR · PRIME (15M · 9–12 ET) — 69.2% WR · PF 1.88 · +$75,990 net · 1.93% max DD · 30 of 30 profitable months (backtest)
• ⚡ RAPTOR · EXTENDED (15M · 9–1 ET) — max signal flow · 68.8% WR · PF 1.78 · +$89,870 net · 1.78% max DD · 11 of 11 profitable months (backtest)
• 🏆 RAPTOR · GOLDEN (15M · 10–1 ET) — live-validated · 88.7% WR · PF 3.13 · 62 verified trades (10am–1pm window) · backtest: PF 1.59 / 69.7% WR · the narrow golden-window subset we trade live

Backtest numbers look modest by design. TradingView's bar magnifier inflates tight-stop strategies by up to 5× — most algo sellers exploit this to publish optimistic-but-fake numbers. We tuned around what survives a real broker (wider stops, cleaner sessions, fewer trades) and validated on a live account.

Live broker results (RAPTOR · GOLDEN tier — 62 verified trades, 10am–1pm window)

Sample week on our personal NinjaTrader account running the RAPTOR · GOLDEN config above:

• $2,582 net across 5 trading days
• 88.7% win rate across 62 trades that week
• 3.13 profit factor across those 62 trades (10am–1pm window)
• Real money: $900 funded → +$2,582 net, ATR risk-based sizing (cap 2), real slippage + exchange fees
• Same RAPTOR · GOLDEN config, backtested: PF 1.59 · 69.7% WR · 2.32% max DD (Jul 2025 – Jun 2026, TradingView Strategy Tester)
• Live PF 3.13 vs its own backtest PF 1.59 — live beats backtest ~2×. Same settings. Empirically rare in this space.

BTC backtest (Elite-exclusive 24/7 config)

On KRAKEN:BTCUSD 30-min, CRYPTO session mode, with the same default settings and 1 contract on $100K capital:

• $207,587 net (+208%) across 5,049 trades over a 9-month window
• vs buy-and-hold: BTC itself fell −36.4% over the same window → the strategy outperformed buy-and-hold by +$244,012 (market-neutral edge, not a BTC-rally bet)
• 64.9% win rate · 1.59 profit factor
• 2.58% max drawdown · 9 of 9 profitable full months (9-month window)
• Best month: Feb 2026 at +$45,959
• ⚠ BTC has NOT been live-validated yet — unlike MNQ RAPTOR · GOLDEN (62 verified live trades in the 10am–1pm window), BTC is backtest only. Apply a heavy real-world haircut and trade demo for 7+ days before going live.

04What the backtest does NOT capture

If we tried to hide this section, you should walk away. Here it is up front.

• Slippage variance. The 1-tick slippage assumption is reasonable for liquid hours on MNQ. During news events, illiquid sessions, or thin crypto pairs, real slippage can be 2-5× larger. This compresses winners and expands losers.
• News halts and limit moves. Major news (FOMC, NFP, geopolitical shocks) can produce gaps the backtest does not model. We disable signals during scheduled high-impact news windows, but unscheduled events can still hit positions.
• Broker-specific fees and funding. Our BTC backtest used Kraken spot (no funding); if you trade crypto perpetuals, exchanges charge funding rates. Different futures brokers charge different per-contract fees. We used a typical retail rate ($1.50/side futures) — verify your broker's actual rate.
• Webhook delivery time. If signals route through middleware (PickMyTrade, TradersPost, etc.), there's a small latency between TradingView alert and broker execution. Usually under 2 seconds, but spikes happen.
• Your discipline. Backtests assume every signal is taken. In live trading, fear, revenge trading, manual overrides, and missed sessions all degrade results. The strategy is rules-based; you have to be too.
• Regime change. The test window covers roughly a year of 2025–2026 market structure. If the broader market shifts into a regime fundamentally different from what was tested (e.g. extended ranging, persistent low-volatility), signal quality may degrade until we can re-tune.
• Survivorship in the test window. We optimized parameters on this specific window and then reported the results. We disclose this — out-of-sample performance is by definition unknown.

05What we promise vs what we don't

06What could go wrong

Risks every prospective subscriber should consider before clicking subscribe:

• Drawdown will happen. Even the best backtested config has a drawdown number. Going through one in live trading feels much worse than reading about it. Size positions so a drawdown 2× the max we report is survivable.
• Strategy degradation. Market microstructure changes. A strategy that worked in 2025 may need re-tuning in 2027. We will communicate when adjustments are made and why.
• Discretionary override risk. The biggest cause of subscriber underperformance is taking signals manually, skipping ones that "look bad," or adding to losers. The system works because every signal is taken — your discretion is the leak.
• Single-strategy concentration. Falcon AI is one strategy on one or two instruments. It is not a diversified portfolio. Treat it as a sleeve within a broader trading approach, not your entire account.
• Prop firm rule risk. Each prop firm has its own rules (consistency, daily loss, drawdown). The strategy is compatible with several firms, but you are responsible for ensuring your specific firm allows automated execution and that the configuration fits their rules. Full risk disclosure →

07How we treat your money

• All billing through Stripe. We never see your card number. You can view, update, or cancel your subscription anytime at /account.
• 14-day Setup Guarantee. Your subscription is charged at signup and backed by our Setup Guarantee: if you set Falcon up correctly and zero signals fire in the first 14 days, your first month is refunded. See the refund policy.
• Monthly subscriptions cancel anytime. No "you must cancel 14 days before billing" gotchas. Cancel today, access continues until your current period ends, no further charges.
• Annual subscriptions. Annual plans are committed for the year per our refund policy, which still includes the 14-day Setup Guarantee on your first month.
• No upsells inside the product. Once you're a subscriber, you get the signals. We don't drip-feed paid coaching, paid signals, or paid courses on top.

08Third-party verification status

Honest answer: no independent backtest audit exists yet. The numbers on this page were generated by us, using TradingView's strategy tester, on our own configurations. You are taking our word that the methodology described above is what we ran.

What you can verify yourself:

• Subscribe, add Falcon AI to your TradingView chart, open Strategy Tester with the same parameters above, and re-run the backtest on your own account. If you get materially different numbers, email us.
• Compare our claimed live signal output to what fires on your chart in real time. Every signal is timestamped on the chart.
• Confirm the LLC is real (Florida Sunbiz · search "Falcon AI LLC").
• Confirm the Stripe billing is real (you'll see "Falcon AI" or "Falcon AI LLC" on your statement).

If you run an independent audit and want to publish it, we will link to it here — flattering or not.

09Hold us accountable

If anything on this page turns out to be wrong, misleading, or incomplete — email us directly and we will fix it.

Direct line: support@thefalconai.com
Subject prefix for fastest reply: [TRANSPARENCY]