• Created a serviceability sidecar deployment as a platform to debug customer clusters without logging into deployments with sensitive customer data.• Worked as part of a 3-man team to create the backend of the Data Subject Request feature from scratch for the Privacy Ops vertical of the LightBeam product. Resulted in customer acquisition increase and wins over larger, more established competitors in the privacy compliance space.• Optimized the Privacy at Partners feature to work at a scale of millions of documents shared to tens of thousands of partners. Achieved a 40x increase in speed.• Assisted the customer success team with monitoring, triaging and fixing issues in customer clusters. Automated the manual review required per cluster resulting in lowering man hours required from 4 hours (to login and check customer clusters) to 30 minutes (reading a report).

• Worked on XtractEdge (previously NIA) ModelOps - simplified and optimized AI/ML model deployment for large-scale solutions.• Worked extensively with Python, Git, Docker, Kubernetes, and briefly with Elasticsearch + Kibana stack.• Customers include Bank of America, Citi Bank and American Express.

• Selected for Honors research program in IIIT under Dr. Jayanthi Sivaswami (Dean of Academics).• Explored various segmentation techniques in Medical Imaging and applied them on cross-sectional lung x-rays.• Applied representation learning to tackle sparsity and scarcity of data in the field of Medical Computer Vision.• Pursued prediction of Idiopathic Pulmonary Fibrosis progression, combining visual data from scans with time-seriesdata from medical readings aiding early detection and diagnosis.

• Explored three dimensional face reconstruction from low-resolution images.• Implemented and trained a prototype neural network, incorporated the Basel Face Model as a prior, using CCTVvideo for facial reconstruction from low-resolution images.

• Introduced Dialogflow Chatbots into the Vitacloud.io pipeline to automate customer interaction and apply deeplearning.• Deployed a backend server routine to backup and store user metrics and data reliably.