●Lighthouse Project - UI Development: Led UI development for Lighthouse using Angular, TypeScript, HTML, CSS, and APIs—a

tool to manage JIRA tickets. Enabled regulatory understanding, raised JIRAs for new asks, and tracked delivery to ensure compliance and

alignment with business goals.

● Business Analyst for Reconciliation: Collaborated with internal teams and DTCC to reconcile datasets via Python scripts. Flagged

mismatches, coordinated fixes, and translated business inputs into developer tasks to improve data accuracy and team synergy.

● Cloud Data Classification & Automation: Helped classify cloud data into on-prem/off-prem using Python and shell scripts. Built

config files, automated audits, and achieved 100% accuracy under tight deadlines.

● Agentic Regulatory System: Created a system with topic-specific compliance agents and smart query routing. Used semantic vector

search on PostgreSQL for better regulatory data handling.

● CI/CD Pipeline for Shuttlerun Pricing Service: Engineered CI/CD for a C# service generating trading prices. Automated

deployments, unit tests, coverage reporting, SonarQube onboarding, Docker image push to Nexus, and production promotion—cutting

manual effort by ~80% and saving 40+ dev hours monthly.

◦ Environment Comparison: Implemented Python code to validate the number of orders and executions between different environments (PROD/DR) in egKDB. Conducted comprehensive comparisons, resulting in a 95% reduction in discrepancies between environments.

◦ Kubernetes Monitoring and Integration: Setup monitoring infrastructure for Kubernetes. Integrated the monitoring system with ITRS(Infrastructure and Technology Services) system, enabling real-time tracking and analysis. Achieved up to 90% uptime monitoring for Kubernetes resources, ensuring proactive issue identification and resolution.

◦ Developed a completely new python code, for services being shifted to the KDB database, and the queries were being run manually for

each client, so reduced the manual effort by 85% by developing a code with a scalable config file input, which helped the team run a single script and know whether the EOD reports are being sent to the client in accordance with the SLA, and the result matches the DR environment, and update the status to OK or critical with respect to the result.