Equity Risk and Return Challenge - Team 3(Syntax Squad)

The Story of PortfolioIQ The inspiration for PortfolioIQ stemmed from the challenge of bridging theoretical mathematical models with the volatile reality of the Indian equity markets. As a Mathematics and Data Science student at HBTU, I wanted to move beyond static analysis and build a "Living Pipeline" for Prayas Capital that could automate the decision-making process for a ₹10 Lakh portfolio. The project was designed to prove that systematic risk management and machine learning can outperform "gut-feeling" investing, especially during periods of high market stress.

The system was built as a modular multi-agent pipeline using Python for the analytical core and n8n for orchestration. I developed a robust data ingestion layer using yfinance, followed by a Strategic Selection engine that isolates the "Elite Two" stocks from the Banking, IT, and Pharma sectors based on their Sharpe Ratios. To add a predictive layer, I engineered a Random Forest model trained on lagged RSI and volatility features, while a low-code automation layer was integrated to provide real-time alerts via Webhooks whenever a "Golden Cross" or significant price movement occurs.

Throughout the process, I learned the critical importance of Data Integrity and the prevention of temporal leakage in time-series forecasting. The project taught me how to transform a local Python script into a production-ready tool that communicates insights directly to a fund manager's dashboard. Understanding that a high-return portfolio is a liability without defensive anchors was a key takeaway, leading to the inclusion of low-beta Pharma stocks to protect capital during the Chaos Round.

The primary challenge was ensuring Feature Alignment—mathematically synchronizing historical indicators with future performance targets without "peeking" into the future. Additionally, balancing the technical depth of seven complex tasks within a strict hackathon deadline required a modular programming approach, allowing individual components like the Risk-Return Map and the n8n workflow to be developed and tested in parallel. This journey has been a masterclass in applying data science to preserve and grow wealth under pressure.

Built With

• n8n
• numpy
• pandas
• python
• scikit-learn
• yfinance