Jun 2025 – Present · 9 mos · Hybrid

• Working on change detection in SAR (Synthetic Aperture Radar) imagery using Transformer-based deep learning models,
• focused on spatiotemporal feature extraction and robustness for ML/AI pipelines.

Jun 2025 – Aug 2025 · 2 mos · Gurugram, Haryana, India · On-site

• Engineered interactive dashboards and automated analytical workflows using Power BI and Alteryx, providing actionable business insights for cross-functional teams. Enhanced automated reporting by integrating GenAI and LangGraph modules via Python.
• Configured Apache Airflow Directed Acyclic Graphs (DAGs) to orchestrate and monitor ETL pipelines, enabling seamless data integration from OLTP to OLAP systems and leveraging BigQuery as the centralized data warehouse for efficient, large-scale analytics; ensured robust data storage with Amazon S3.
• Automated financial reporting with runtime query generation and batch computation using Google BigQuery and GCP, improving processing scalability and reducing report latency by 45% and applied advanced statistical and financial modeling techniques to support forecasting and strategic decision-making.

Jun 2025 – Aug 2025 · 2 mos

Jan 2025 – Mar 2025 · 2 mos · Bengaluru, Karnataka, India · On-site

• Improved scalability of Trellix’s Cairo-UI by developing reusable Angular components and integrating backend services via Node.js, enhancing performance and reducing dev overhead.
• Strengthened product reliability by implementing unit test suites using Jasmine and Karma, increasing test coverage by 30% and significantly reducing UI regressions across production releases.

Jul 2024 – Aug 2024 · 1 mo · Delhi, India · On-site

• Led research on adversarial voicebox attacks using CNNs trained on CIFAR-10 and CIFAR-100 datasets, achieving over 85% attack success rate using black-box methods such as SimBA, Square, and NES attacks.
• Developed and fine-tuned custom convolutional neural networks (CNNs), improving defense mechanism efficiency by 20% through adversarial training and optimization techniques tailored to specific models.
• Authored in-depth analytical reports identifying key vulnerabilities in CNN models against adversarial perturbations, directly influencing a targeted defense strategy that enhanced model robustness by 20%, and prioritized solutions to mitigate high-severity attacks.

Dec 2023 – Jan 2024 · 1 mo · New Delhi, Delhi, India · On-site

• Streamlined CNS operations by contributing to automation using systems like AMSS and AMHS, reducing manual workload and enabling faster flight message processing across AAI teams.
• Assisted in radar, S-band and VHF system checks while working with SAP ERP modules, cloud platforms, and database management systems, improving system reliability and aiding smoother real-time communication across
• departments.