AI Trading Bot with SMC for XAU/USD - Video Script
Title: Build an AI Trading Bot using Smart Money Concepts (SMC) for XAU/USD in Python
[Intro Slide | 0:00 - 0:30]
"Welcome to this deep-dive tutorial on building an AI trading bot using Smart Money Concepts (SMC) for
XAU/USD, also known as gold versus the US dollar. If you're into forex or commodities trading and want
to combine institutional trading techniques with machine learning, you're in the right place."
[Slide: What You'll Learn | 0:30 - 1:00]
- How to get live gold data in Python
- Detect Smart Money structure: HH, LL, BOS, OB, FVG
- Train a machine learning model to predict trades
- Backtest the strategy
- Build your own bot logic
"By the end, you'll have a working foundation of a strategy that blends the power of price action with
predictive AI models."
[Section 1: Setup & Data Collection | 1:00 - 2:30]
Voice-over:
"First, we install the required libraries using pip. These will let us collect gold price data, visualize it, and
run our AI models."
Code:
pip install yfinance pandas numpy matplotlib scikit-learn ta
"Next, we fetch hourly XAU/USD data using Yahoo Finance."
import yfinance as yf
data = yf.download("XAUUSD=X", start="2023-01-01", end="2025-07-01", interval="1h")
data = data[['Open', 'High', 'Low', 'Close', 'Volume']].reset_index()
AI Trading Bot with SMC for XAU/USD - Video Script
[Section 2: Market Structure (HH, LL, BOS) | 2:30 - 5:00]
Voice-over:
"Now we detect swing points: higher highs and lower lows. Then, we'll track breaks of structure (BOS)."
Code:
def detect_swings(df):
df['HH'] = (df['High'] > df['High'].shift(1)) & (df['High'] > df['High'].shift(-1))
df['LL'] = (df['Low'] < df['Low'].shift(1)) & (df['Low'] < df['Low'].shift(-1))
return df
def detect_bos(df):
trend = None
last_hh = last_ll = None
bos = []
for i in range(len(df)):
row = df.iloc[i]
if row['HH']: last_hh = row['High']
if row['LL']: last_ll = row['Low']
if last_hh and row['Close'] > last_hh and trend != 'bullish':
bos.append('BOS UP')
trend = 'bullish'
elif last_ll and row['Close'] < last_ll and trend != 'bearish':
bos.append('BOS DOWN')
trend = 'bearish'
else:
bos.append("")
df['BOS'] = bos
return df
# Apply it
data = detect_swings(data)
data = detect_bos(data)
AI Trading Bot with SMC for XAU/USD - Video Script
[Section 3: Order Block Detection | 5:00 - 7:00]
Voice-over:
"Order blocks are key SMC zones where institutions place orders. We'll find the last opposite candle
before a BOS."
Code:
def detect_order_blocks(df):
bullish_obs, bearish_obs = [], []
for i in range(2, len(df)):
if df['BOS'].iloc[i] == "BOS UP":
for j in range(i-1, i-4, -1):
if df['Open'].iloc[j] > df['Close'].iloc[j]:
bullish_obs.append(df.iloc[j])
break
elif df['BOS'].iloc[i] == "BOS DOWN":
for j in range(i-1, i-4, -1):
if df['Open'].iloc[j] < df['Close'].iloc[j]:
bearish_obs.append(df.iloc[j])
break
return pd.DataFrame(bullish_obs), pd.DataFrame(bearish_obs)
bullish_ob_df, bearish_ob_df = detect_order_blocks(data)
[Section 4: Fair Value Gaps (FVG) | 7:00 - 9:00]
Voice-over:
"Next up are Fair Value Gaps or FVGs, which are imbalances that price often returns to fill."
Code:
def detect_fvgs(df):
bullish_fvgs, bearish_fvgs = [], []
for i in range(2, len(df)):
AI Trading Bot with SMC for XAU/USD - Video Script
prev = df.iloc[i-2]
curr = df.iloc[i]
if curr['Low'] > prev['High']:
bullish_fvgs.append({'Datetime': curr['Datetime'], 'GapLow': prev['High'], 'GapHigh': curr['Low']})
if curr['High'] < prev['Low']:
bearish_fvgs.append({'Datetime': curr['Datetime'], 'GapLow': curr['High'], 'GapHigh':
prev['Low']})
return pd.DataFrame(bullish_fvgs), pd.DataFrame(bearish_fvgs)
bullish_fvg_df, bearish_fvg_df = detect_fvgs(data)
[Section 5: Machine Learning Prediction | 9:00 - 12:00]
Voice-over:
"We now use these features to train a machine learning model that predicts the next price movement."
Code:
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
# Example feature set
data['SMA50'] = data['Close'].rolling(window=50).mean()
data['SMA200'] = data['Close'].rolling(window=200).mean()
data.dropna(inplace=True)
features = data[['Close', 'SMA50', 'SMA200']]
labels = data['Close'].shift(-1)
X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.2, shuffle=False)
model = RandomForestRegressor()
model.fit(X_train, y_train)
AI Trading Bot with SMC for XAU/USD - Video Script
predictions = model.predict(X_test)
print("MSE:", mean_squared_error(y_test, predictions))
[Section 6: Create Trade Signal | 12:00 - 13:30]
Voice-over:
"Finally, we generate a trade signal based on the AI model's prediction and combine it with SMC context."
Code:
latest_features = features.iloc[-1].values.reshape(1, -1)
next_price = model.predict(latest_features)[0]
current_price = data['Close'].iloc[-1]
if next_price > current_price:
print("AI Signal: BUY")
elif next_price < current_price:
print("AI Signal: SELL")
else:
print("AI Signal: HOLD")
[Outro | 13:30 - 14:00]
"And that’s it! You’ve now built a Smart Money Concept trading bot foundation in Python for XAU/USD.
From order blocks to fair value gaps to machine learning — all in one system."
"If you found this valuable, leave a like and comment below if you want the full deployment to
MetaTrader or TradingView via webhook."
