Built an Agentic AI product delivering hallucination-free, source-aware responses by integrating enterprise databases, enabling dynamic data modeling, and powering context-rich queries with document uploads, agent transparency, and tool integrations (e.g., web search). Modernized multilingual support by replacing legacy translation with LLM-powered pipelines, enabling seamless, high-accuracy global usage.

Drove PSR (Performance, Scalability & Reliability) testing across releases—designed test plans, defined SLAs, and executed stress tests—achieving a 50% increase in concurrent capacity and 35% faster indexing.

Course: Software Project Management

As part of the Model Risk Governance and Review (MRGR) Ongoing Performance Monitoring team, I was working on the technology platforms that assesses performance of quantitative models used throughout the firm.

• Developed and deployed a web-based analytical tool using Python3, JavaScript, and Tornado, serving over 500 users. This tool significantly improved real-time analytics and streamlined data processing for risk assessment models.

• Led a team of 15 to deliver four key projects focused on report automation and enhanced dashboard functionalities. Developed APIs to identify and highlight critical insights for senior risk management, improving decision-making processes.

• Automated the dispatch of 10,000+ reminder emails annually through a Python-based job, significantly reducing manual tasks and enhancing operational efficiency.

• Implemented key performance metrics in a centralized analytics application to identify and address underperforming models. Utilized heat map analysis for effective feature selection, leading to more accurate and robust model performance evaluations.