Data Science & AI enthusiast — building end-to-end ML systems, researching medical image segmentation

github

Shiv Narain  • Peerlist

• Designed a machine learning pipeline using TimeSeriesSplit and 14 technical indicators to generate next-bar trading signals with confidence scoring • Implemented a vectorized backtesting engine modeling transaction costs, slippage, and risk metrics for realistic performance evaluation • Built an interactive dashboard and modular CLI integrating Binance API, enabling real-time signal monitoring and controlled order execution

AlphaPulse

• Developed an AI-based application to summarize documents, extract key insights, and enable question-answering using LLM APIs • Implemented document parsing, prompt engineering, and chunk-based processing for handling inputs • Built a Streamlit interface for real-time interaction and automated document analysis

AnyDoc

• Developed a URL shortening service with unique link generation, redirection, and click tracking, handling high-frequency requests with integrated analytics • Integrated Redis caching to optimize lookup performance and reduce database load, improving efficiency