I need a compact, reliable real-time candlestick pattern recognition engine that integrates with my existing market data feed. The streaming API is already available and delivers live OHLCV data in JSON format, so the work is focused entirely on pattern r
Budget: $250 – $750 USD
I need a compact, reliable real-time candlestick pattern recognition engine that integrates with my existing market data feed. The streaming API is already available and delivers live OHLCV data in JSON format, so the work is focused entirely on pattern recognition logic, not data acquisition.
The system must continuously scan incoming candlestick data and detect three specific chart patterns: Head and Shoulders, Double Top/Bottom, and Cup and Handle. When a pattern is completed or clearly in its final confirmation stage, the engine should trigger an instant alert that can be routed to a webhook, email, or simple on-screen notification.
A clean, well-documented Python module using pandas, NumPy, or TA-Lib is preferred, provided it processes live ticks with a maximum end-to-end delay of two seconds. The solution must correctly identify at least 90% of supplied test patterns and emit a single structured alert per event, including timestamp, symbol, pattern name, and confidence score. Concise documentation and a short demo script running against my sandbox feed are required. Optional enhancements such as adjustable sensitivity thresholds or a lightweight chart overlay are welcome, but the core deliverable is the real-time detection engine.
⸻
If you want, I can also:
• Make it more technical
• Make it more beginner-friendly
• Tailor it for Upwork, Freelancer, or Toptal
The system must continuously scan incoming candlestick data and detect three specific chart patterns: Head and Shoulders, Double Top/Bottom, and Cup and Handle. When a pattern is completed or clearly in its final confirmation stage, the engine should trigger an instant alert that can be routed to a webhook, email, or simple on-screen notification.
A clean, well-documented Python module using pandas, NumPy, or TA-Lib is preferred, provided it processes live ticks with a maximum end-to-end delay of two seconds. The solution must correctly identify at least 90% of supplied test patterns and emit a single structured alert per event, including timestamp, symbol, pattern name, and confidence score. Concise documentation and a short demo script running against my sandbox feed are required. Optional enhancements such as adjustable sensitivity thresholds or a lightweight chart overlay are welcome, but the core deliverable is the real-time detection engine.
⸻
If you want, I can also:
• Make it more technical
• Make it more beginner-friendly
• Tailor it for Upwork, Freelancer, or Toptal