AgentSmyth is a leader in financial analysis automation. AgentSmyth provides a suite of smart autonomous agents that act as specialized team members, revolutionizing coordination in finance. This includes macro research, expertise at both the ticker and sector levels, and monitoring of unusual option flow analysis, all while delivering tick-by-tick narratives.

MarketReader is a market intelligence company that empowers all investors to understand why the market is moving in real time. MarketReader aims to provide clear, objective explanations of asset price movement in a way that is accessible to all types of investors, from retail to institutional, using our proprietary technologies for transforming market data, obtaining relevant 3rd party data, and combining these varied sources of data as input for an LLM (Large Language Model) to provide concise, timely, powerful yet easily-accessible explanations.

Designed, optimized and managed extensive data pipelines that combined human genetics, gene expression, and other demographic data (daily processing of Terabytes of data). The pipeline was executed in Batch Computing environment in Princeton University cluster that involved ETL, CPU and memory management, speed optimization with vectorization, and iterative ML pipelines that incorporate previous findings.Through these enhancements, I scaled the extensive ML pipeline for over 1000 donors (Terabytes of data) from days to less than three hours. My significant contributions, both technical and scientific, led to a first-author recognition in a Nature publication for an NIH consortium.

▪ Utilizing Apache Hadoop, Apache Solr and Data Fusion technology to develop User Behavior Analysis and Recommendation Engines based on User-Generated Big Data▪ Developing / Maintaining vital applications for E*Trade Retail Brokerage Platform: Enterprise search engine▪ Enterprise Internal Hackathon winner: Natural Language Processing (NLP) functionality for Stock Screener

▪ Scientific Publication: Paul Nerenberg, BRIAN JO, Clare So, Ajay Tripathy, Teresa Head-Gordon (2012) Optimizing Solute-Water van der Waals Interactions To Reproduce Solvation Free Energies.▪ Technology Presentation: Berkeley Mini Stat-Mech Conference, January, 2012. ▪ Highlights: Participated in the entirely computational research conducted in UNIX environment. Learned the Shell scripting languages and was introduced to Perl, managed a huge database of information, and successfully engineered an optimization method for protein folding simulation calculation algorithms to reduce RMS error by 65% and mean signed error by 85% which led to a scientific publication.

Evaluated the significance of Network analysis of genetic regulation and patterns of gene expression through statistical analysis of approx. 100,000 human lung adenocarcinoma genes and matched normal genes, conducted in Visual Basic and targeted mutation segments that show statistical significance.