Credit Systematic Market Making- Researched on Corporate Bond Liquidity: Validated a model which utilizes past volume, bond characteristic features combining into a score; compared it against MarketAxess Liquidity Scores on TRACE, and other extensive bond volume datasets fetched from KDB+; Productionized the liquidity score validation framework using Python- Improved the model by 3% by enhancing one of its key features, leveraging insights from Machine Learning techniques Elastic Net, Random Forest, and XGBoost for feature selection and identifying non-linear relationships

Designed a machine learning framework integrating aggregation, smoothing, and rigorous testing pipelines to predict macroeconomic indicators (GDP, inflation, consumer index), leveraging survey data from AlphaROC's Occam

Knowledge Graph Team- Designed and Maintained big data solutions on top of peta-scale, firm-wide data modeled into a graph with vertices (entities) and edges (relationships) for applications to banking activities like compliance and investment banking- Developed the infrastructure for a knowledge graph deduplication system from scratch in Hadoop MapReduce, sourcing data from multiple sources (60 million vertices, 4 billion edges), leading to duplication of nodes in the graph- Designed an algorithm using the longest common subsequence to identify the company domain and then the best person in graph for making introductions to potential clients of IBD; Integrated using REST API, Spring in Java

Research in Financial Markets on Low Frequency OHLCV data- Designed and implemented alpha-generating signals in C++ for optimizing weight allocation across a portfolio of 130 Indian stocks, utilizing Technical Indicators on OHLCV daily frequency data, Sharpe ratio for performance evaluation- Leveraged Time Series forecasting techniques ARCH, and GARCH to estimate volatility as a risk measure in alphas- Conducted comparative analysis of Deep Learning models (RNN, LSTM, CNN) for predicting stock prices