A complete walkthrough of every feature, the technology behind the platform, and how to get the most out of AI-powered pattern recognition.
StockTrendz AI™ is an intelligent trading analytics platform that uses machine learning to detect, score, and explain chart patterns in real time — giving every trader an AI co-pilot.
The platform continuously scans live candlestick data and compares it against a large historical database using multiple similarity algorithms. When a strong pattern match is found, you get instant visual feedback and an AI-generated insight.
Real-time candlestick data combined with historical analog matching — all in one powerful view.
The top row of stat cards gives you an immediate snapshot: current asset price, the pattern similarity score for the best historical match, active alert count, and the last-refresh timestamp.
The right panel displays the closest historical window found — showing the full candlestick sequence side-by-side with similarity metrics beneath the chart.
Search any NSE/BSE stock and receive an AI-scored Bullish Signal Detected / Hold / Sell signal backed by live pattern engine data and fundamental company information.
Enter a ticker (e.g. TCS, INFY, RELIANCE) and the platform instantly scores the stock on a 0-100 composite scale, combining technical pattern signals with momentum indicators.
An interactive charting workspace that overlays real-time candlestick data with AI-detected pattern matches and probability rankings from a built-in pattern library.
The right sidebar continuously evaluates the current chart against a pattern library of 30+ classical formations, scoring each one and ranking them by probability. An Overall Sentiment gauge combines all signals into a Bullish / Bearish / Neutral reading.
From raw market data to actionable AI insights — four tightly integrated steps run every time you load a chart.
The backend connects to live market APIs (Yahoo Finance / NSE feed) and streams OHLCV candlestick data into the platform. Each candle is validated, normalised, and stored for pattern comparison.
The AI engine extracts the most recent N candles, normalises the price sequence, and runs it through DTW, Cosine Similarity, and Pearson Correlation algorithms against a large historical dataset to find the closest analogs.
The best historical match window is rendered as a full interactive chart alongside the live chart. Similarity metrics (DTW, Cosine, Pearson, Movement) and technical indicators (RSI, SMA-20, SMA-50) are displayed below for comparison.
The AI layer synthesises all signals into natural-language insights: trend direction, RSI status, pattern match confidence, and next-candle probability estimates. Alerts can be set to trigger when conditions are met.
Every layer of the stack is chosen for performance, reliability, and developer velocity.
The server-side engine is built with Flask, handling API routing, data ingestion, and the pattern recognition algorithms. NumPy and SciPy power the similarity computations.
Historical OHLCV data, user accounts, alert configurations, and pattern metadata are stored in a robust PostgreSQL database with time-series optimisations.
Interactive candlestick charts are rendered client-side using TradingView's Lightweight Charts library, delivering fast, smooth 60fps performance.
A custom similarity engine uses Dynamic Time Warping (DTW) to compare normalised price series, with Cosine and Pearson correlation as secondary scoring dimensions.
Live and historical data is sourced via a REST API layer that wraps Yahoo Finance and direct NSE/BSE feeds, with automatic retry and caching logic built in.
The frontend is handcrafted in vanilla HTML, CSS, and JavaScript — no heavy framework overhead — resulting in fast page loads and a smooth user experience.
Quick definitions for every metric and term you'll encounter on the platform.
Everything you need to know before getting started.
Join StockTrendz AI™ and let AI find your next high-confidence trade.