Seven separate readings, one number. This confluence indicator TradingView script scores trend, momentum, volume, volatility, MACD, SuperTrend and structure from zero to one hundred each. It then blends them with weights of 25, 20, 15, 10, 15, 10 and 5 percent. The blend produces a bull score and a bear score. A marker prints when either one crosses your confidence threshold, 70 by default, on a closed bar.
It is free and open source. You can add it to any chart from the AI Signal Confluence script page on TradingView, and the full Pine source is published there for you to read.
What AI Signal Confluence plots on the price chart
The script overlays the price chart. Every bar takes a gradient color that runs from pink at a bull score of zero to cyan at one hundred. The background carries the same gradient at 95 percent transparency. Signal bars get a three-layer triangle – huge, large, then solid – below the bar for buys and above it for sells. A label shows BUY or SELL with the score to one decimal. A ten-row dashboard lists each component, its weight, a colored vote dot, its value, a block confidence bar and a final verdict.

How AI Signal Confluence is built
Seven component scores
Trend reads three EMAs. A full 9 over 21 over 50 stack scores 100. One aligned pair scores 65. A 9 above the 50 alone scores 40. A full bearish stack scores 0, and anything else scores 30. Momentum grades RSI(14). Readings of 70 and above score 100, and 60 and above score 75. The 40 to 60 band scales up to 50. The low 30s score 15, and below 30 scores 0.
Volume divides current volume by its 20-bar average. A ratio of 2.0 scores 100, 1.5 scores 75, and 1.0 scores 50. Thinner bars score the ratio times 50. Volatility uses Bollinger width on a 20-period, 2.0 deviation band, ranked against the last 100 bars. An expanding width scores that percentile plus 20, capped at 100. A quiet squeeze under the 20th percentile scores 30.
MACD normalises the 12/26/9 histogram against its highest absolute value over 50 bars, then scales that to 100. SuperTrend with length 10 and factor 3.0 votes a flat 100 or 0. Structure places the close between the last confirmed pivot low and pivot high, using a five-bar pivot on each side, and reports that position as a percentage.
Blending and gating
The engine multiplies each component score by its weight, sums them, then divides by the total weight. The defaults add to 100, yet the code divides by whatever your seven weights total. So you can push Trend to 60 without breaking the scale. Both scores then clamp between 0 and 100.
Two components stay direction-neutral on purpose. Volume and volatility feed the same value into the bull and bear sums, because a busy or expanding market helps neither side by itself. The other five invert, so the bear score mirrors the bull score for trend, momentum, MACD, SuperTrend and structure.
A signal needs a fresh cross. The bull score must cross above the threshold on that bar, the bar must be confirmed, and more than 15 bars must have passed since the previous signal. The confirmed-bar rule means the script does not repaint. Despite the name, no machine learning runs here. The code holds no model, no training step and no neural network – only fixed rules and weights you set.
How to read the signals
Read the vote column before the headline number. Seven dots turn green, grey or pink. A confidence of 72 driven by trend, MACD and SuperTrend describes a different market from a 72 propped up by volume and volatility, which pick no side at all.
The verdict row shows the stronger of the two scores with an arrow. Bar color tells the same story continuously. So a chart drifting from pink to cyan across ten bars shows the components flipping one by one. Markers appear only on the bar where the score first crosses the threshold.
Tune with the weights, not just the threshold. Range traders can cut Trend to 10 and raise Momentum and Structure. Trend followers can do the opposite. Raising the threshold from 70 to 85 cuts marker count sharply, because the weighted average rarely reaches the high 80s without near-total agreement.

Every AI Signal Confluence setting explained
The script exposes 23 inputs, grouped in the settings panel exactly as shown below. Defaults are the published values.
Signals
| Setting | What it does | Default | Range |
|---|---|---|---|
| Confidence Threshold (%) | Minimum confidence to trigger a signal | 70 | 50 to 95 |
| Signal Cooldown (bars) | Minimum bars between signals | 15 |
Indicator Settings
| Setting | What it does | Default | Range |
|---|---|---|---|
| EMA Fast | Sets the ema fast used in the calculation. | 9 | |
| EMA Mid | Sets the ema mid used in the calculation. | 21 | |
| EMA Slow | Sets the ema slow used in the calculation. | 50 | |
| RSI Length | Sets the rsi length used in the calculation. | 14 | |
| BB Length | Sets the bb length used in the calculation. | 20 | |
| BB Multiplier | Sets the bb multiplier used in the calculation. | 2.0 | |
| MACD Fast | Sets the macd fast used in the calculation. | 12 | |
| MACD Slow | Sets the macd slow used in the calculation. | 26 | |
| MACD Signal | Sets the macd signal used in the calculation. | 9 | |
| SuperTrend Length | Sets the supertrend length used in the calculation. | 10 | |
| SuperTrend Factor | Sets the supertrend factor used in the calculation. | 3.0 | |
| Pivot Lookback | Sets the pivot lookback used in the calculation. | 5 |
Advanced: Weights
| Setting | What it does | Default | Range |
|---|---|---|---|
| Trend Weight % | Sets the trend weight % used in the calculation. | 25 | 0 to 100 |
| Momentum Weight % | Sets the momentum weight % used in the calculation. | 20 | 0 to 100 |
| Volume Weight % | Sets the volume weight % used in the calculation. | 15 | 0 to 100 |
| Volatility Weight % | Sets the volatility weight % used in the calculation. | 10 | 0 to 100 |
| MACD Weight % | Sets the macd weight % used in the calculation. | 15 | 0 to 100 |
| SuperTrend Weight % | Sets the supertrend weight % used in the calculation. | 10 | 0 to 100 |
| Structure Weight % | Sets the structure weight % used in the calculation. | 5 | 0 to 100 |
Display
| Setting | What it does | Default | Range |
|---|---|---|---|
| Show Dashboard | Toggles show dashboard on the chart. | on | on / off |
| Show Confidence Background | Toggles show confidence background on the chart. | on | on / off |

Alerts built into AI Signal Confluence
The script ships 11 alert conditions. Open the alert dialog on the chart, pick the indicator as the condition source, then choose the event you want. Alerts fire on the close of the bar, so they follow the same confirmed-bar rule the on-chart signals use.
- AISC Buy Signal
- AISC Sell Signal
- Any AISC Signal
- High Confidence (>80%)
- Extreme Confidence (>90%)
- Trend Fully Aligned
- Volume Spike (>2x)
- All Indicators Agree
- Webhook JSON
Other markets and timeframes
Every component works in relative terms. RSI, percentile-ranked Bollinger width, a volume ratio against its own average, and a MACD histogram normalised by its own 50-bar extreme all rescale automatically. The script quotes no fixed pip or dollar values, so it runs on forex, indices, metals, stocks and crypto unchanged. Volume is the one weak spot. Forex tick volume works as an activity proxy, and symbols with no volume feed push that 15 percent weight toward zero.

Limitations worth knowing
The name says AI, and the code does not. Seven fixed formulas produce seven scores, then static weights combine them. Nothing trains, adapts or learns from past signals. Treat the confidence number as a weighted vote count, not a forecast.
Two of the seven inputs never choose a direction. Volume and volatility add the same points to both sides, so 25 percent of the default weight lifts the bull and bear scores together. That inflates both readings in busy markets and can push a thin setup across the threshold.
The tool is entry-only. It plots no stop, no target and no position sizing. It also needs a fresh threshold cross, so a score that sits at 85 for thirty bars produces exactly one marker. Long trends can therefore show far fewer arrows than the background color suggests.
Get AI Signal Confluence on TradingView
Open AI Signal Confluence on TradingView
If you also trade MetaTrader, the MT4 and MT5 indicator library is available below.
Get the complete indicator library
One email unlocks the full MT4 and MT5 indicator library. This TradingView script stays free on TradingView – the button above adds it to your chart.
Using it alongside MetaTrader
Adding a script on TradingView takes one click, so there is no install step here. If you want the same idea on MetaTrader, the MT4 and MT5 indicator installation guide walks through copying files into the data folder and attaching them to a chart. You can also browse the full MetaTrader indicator library, the MT4 indicators section, or the other free TradingView scripts published on this profile. For related chart tools see the MT5 indicators section and the forex trading strategies guides.
External references
- Technical analysis on Wikipedia
- Momentum at Investopedia
- Advance Decline Indicators at StockCharts ChartSchool
Frequently asked questions
Is AI Signal Confluence actually machine learning?
No. The script runs seven classic calculations – EMA stack, RSI, relative volume, Bollinger width percentile, MACD histogram, SuperTrend and pivot structure – then combines them with fixed percentage weights. The code holds no model, no training data and no neural network.
Does it repaint?
No. Both signal conditions require barstate.isconfirmed, and alerts use alert.freq_once_per_bar_close. Scores update live inside the bar, but markers print only once a bar closes.
How do I make it fire more often?
Lower the Confidence Threshold toward 60 and shorten the cooldown from 15 bars. You can also rebalance the weights. Dropping SuperTrend and Structure to zero puts the whole decision on trend, momentum, volume, volatility and MACD.
Do the weights have to add up to 100?
No. The code divides the weighted sum by the total of your seven weights, so any set rescales correctly. Setting all seven to zero falls back to a divisor of 100. Trading carries risk, and past performance does not indicate future results.
How much should I rely on this indicator?
Treat it as one input, not a decision. It is a chart analysis tool, not trading advice. Test it on your own markets and timeframes before relying on it. Results are not guaranteed; past performance is not indicative of future results.
