Member of Technical Staff 

Software Engineer Intern

• Designed and implemented a multi-agent Al system (LangGraph + OpenAI APIs) to automate AML case investigations across ADP's Financial Crime Risk Management (FCRM) platform, supporting ~90,000 cases annually.
• Built REST-driven backend services to aggregate AML Investigation data across USMM, IMM, and WM systems, retrieving transaction details and prior case history from multiple production databases, and structured the consolidated data (via JSON/Pandas processing) as contextual inputs for LLM-based risk assessment workflows.
• Reduced manual case processing time from 25-30 minutes to -2 minutes per case (~80% efficiency gain) by generating structured case descriptions, dynamic risk scores, and regulator-ready investigative narratives.
• Optimised SQL queries on large Oracle production databases (10M+ records), improving query performance for Oracle FCCM workflows, and integrated into the FCRM Power Bl dashboard used by 50+ business and compliance stakeholders across India & the US.
• Enhanced Unix-based ETL workflows by automating ingestion of IMM, USMM, Work Market (WM), and Recon Master datasets into production databases, improving reporting accuracy and data reliability. 

Machine Learning Research Intern 

• Worked under Dr. Arif Ahmed Sekh 

• Developed FairMed - XGB, a fairness-aware machine learning framework to reduce gender bias in critical-care prediction models while preserving clinical performance.
• Designed a multi-metric fairness optimisation pipeline using Bayesian optimisation, balancing accuracy with equity metrics such as Statistical Parity Difference, Theil Index, and Wasserstein Distance.
• Applied the framework to real-world ICU datasets (MIMIC-IV, eICU), achieving a substantial reduction in gender-based prediction disparities across patient cohorts.
• Integrated model interpretability using SHAP, enabling clinicians to identify how specific features contribute to biased predictions and supporting transparent, trust-aware clinical decision-making. 

Software Engineer Intern 

• Worked under Dr. Sourav Kaity 
• Built a real-time radar-tracking simulator in Java, processing live trajectories of airborne objects and visualising key parameters to support system analysis and testing.
• Developed a low-latency client–server communication system using UDP sockets, enabling concurrent data transmission, reception, and real-time visualisation across networked systems.
• Designed modular components for data ingestion, coordinate transformation, and multi-threaded processing, ensuring scalable and maintainable system architecture.
• Developed interactive visualisation interfaces using GLG Toolkit and web technologies, allowing real-time monitoring, debugging, and analysis of radar tracking data. 

BETA - Student Ambassador

AI/ML Lead 

Reviewer 

Best Paper Award

• Worked with Dr. Somnath Mahato on the paper titled "Advancing Adversarial Detection of Anomalous Wi-Fi Activity Through Residual Generative Adversarial Networks: A Novel Approach for Robust Wireless Network Security"
• Presented the paper and stood out based on the joint scores of Reviewers + Session Chairs

Letter of Appreciation

• Recognized by NASA’s Vulnerability Disclosure Program for responsibly identifying and reporting a critical security vulnerability as an independent researcher. 
• Commended by NASA OCIO and SAISO for enhancing cybersecurity awareness and contributing to the protection of NASA’s information infrastructure.