Скачать или смотреть I Built an ML-Assisted Intraday Trading Bot in Python

I built a Python intraday trading bot for NIFTY — but it’s not just a trading experiment.
It’s a complete data-to-decision system, built to explore how data engineering, logic, and a bit of ML can come together in one real-time pipeline.

As a Senior Data Engineer, this project was my way to learn how real-time data systems evolve into AI-driven architectures.

What the bot does:
Detects Support & Resistance zones using DBSCAN clustering (unsupervised ML)
Integrates VWAP and Volume Profile zones for volume-aware decisions
Analyzes candlestick patterns and trend direction
Applies volatility, volume, and distance filters
Generates confidence-scored signals for backtesting and evaluation

Architecture Overview:
Market Data → Support/Resistance Engine (DBSCAN + VWAP + Volume Profile) → Signal Engine (Pattern + Trend + Filters) → Backtesting & Risk Manager

Tech stack:
Python, pandas, numpy, sklearn, TA-Lib, Angel SmartAPI

Goal:
Not to predict the market — but to understand how **data pipelines can evolve into intelligent, explainable decision systems**.

---

GitHub Repository: https://github.com/abhishekgopalaiah/...

---

Chapters:
00:00 – Intro
00:13 – Why I built this
00:42 – Architecture Overview
01:45 – Project Structure
03:30 – Code Overview
08:00 – Backtesting and log sample
11:58 – Engineering Reflections and Closing Thoughts

About Me:
I’m Abhishek — a Senior Data Engineer exploring how data systems can make intelligent decisions.

#DataEngineering #AI #Python #TradingBot #MachineLearning #Quant #Architecture #TechJaala #NIFTY #DBSCAN