Jul 2021 – Present · 4 yrs 10 mos · On-site

• Designed ,developed and deployed a real-time Level-0 data quality system for thermal Earth observation satellites, enabling early anomaly detection in blackbody and deep space data.
• Designed and implemented an automated cross-calibration framework for multi-sensor optical satellite imagery aligning it with reference sensors (Landsat-8, Sentinel-2A, MODIS) across multiple data formats.
• Engineered a computer vision pipeline to accurately quantify geo-location discrepancies between Indian Metrological Satellite and Landsat-8 satellite imagery, improving alignment precision by 20%. Presented this work at InGARSS 2024 -Goa (IEEE Indian Geoscience and Remote Sensing Symposium) showcasing practical applications for real-time Earth observation systems.

Jan 2020 – Jun 2021 · 1 yr 5 mos · On-site

• Designed, developed, and deployed a Real-Time Level-0 data quality evaluation system for Microwave Earth observation satellites , processing raw microwave imaging data to generate automated health metrics for early fault detection and system reliability assurance.
• Engineered a radiometric cross-calibration system to validate Level-1 imagery from high-resolution Indian satellites against comparable products from NASA (Landsat) and ESA (Sentinel-2); improved radiometric consistency by 30%, enhancing interoperability across multi-agency datasets.
• Built an end-to-end cloud detection pipeline for grayscale satellite imagery using a U-Net model and a custom-labeled dataset. Automated ground truth generation with superpixel segmentation and adaptive thresholding, followed by visual QA. Presented this work at IGARSS 2021 -Belgium (IEEE International Geoscience and Remote Sensing Symposium).

Jun 2017 – Jan 2020 · 2 yrs 7 mos · On-site

• Led QA initiatives for onboard and ground control software across LEO, GEO, and interplanetary mission modules; ensured system integrity and reliability through semi-automated verification workflows.
• Designed and implemented software simulation models for spacecraft systems using C++ and Python, enabling real-time validation for mission-critical subsystems and significantly reducing test cycle durations by 30%.
• Developed an intra-satellite command encoding/decoding tool, improving test coverage for onboard-ground communication protocols. Improved error handing and encoding/decoding reliability by 50% and contributing to a 30% reduction in ground testing cycles and caught 100+ edge-case failures during pre-launch simulations.
• Built an unsupervised anomaly detection system using clustering techniques (e.g., DBSCAN, K-Means) to identify deviations in spacecraft telemetry; enabled early prediction of onboard system faults.