Ever wondered what happens in those 10 seconds between uploading a chart and getting your analysis? Here's what the AI actually sees.
When you upload a chart to FXSynapse AI, the engine doesn't just look for textbook patterns. It analyzes the entire price structure — candle bodies, wick lengths, spacing, volume context, and multi-layer confluence. Think of it as having a trader with 10,000 hours of screen time reviewing your chart in seconds.
The AI first processes your screenshot to identify the chart area, strip out UI elements, and normalize the image. It detects the chart type (candlestick, line, bar), identifies the color scheme, and maps pixel positions to approximate price levels. This is why cleaner screenshots give better results.
Next, the AI maps out market structure — higher highs, lower lows, ranges, and transitions. It identifies the current trend on the visible timeframe and marks key structural levels that price has respected multiple times.
Support and resistance levels are found by analyzing where price has previously bounced, consolidated, or shown strong reactions. The AI weighs levels by how many times they've been tested, how recently, and how strong the reaction was. These become your drawn levels.
Candlestick patterns (engulfing, hammers, dojis), chart patterns (flags, triangles, head & shoulders), and confluence signals are all identified. The AI doesn't just find patterns — it weighs whether they're in a meaningful location (at support/resistance) or just noise.
Finally, all the data converges into a trade setup: optimal entry, stop loss based on recent swing structure, and take profit targets at the next significant levels. The confidence score reflects how many factors align — more confluence = higher confidence.