• Collaborated with Infineon and T ̈UV S ̈UD to reduce traffic accidents caused bye-scooters.• Conducting user research methodologies and experiments at different stages ofproduct development while ensuring agile development principles.• Developed a Driver Behaviour monitoring and modelling algorithm based on sen-sor data collected from e-scooters. Created a mobile application which providesdriver a score on speed, stability and cautiousness of the ride.

• Trained and optimized deep learning models for image classification and objectdetection to run on ATMs, retail stores, construction sites and production facilities with upto 90% accuracy on edge devices.• Developed and maintained a scalable real-time and post-mortem video analyticsengine integrating open source softwares like OpenCV, Tensorflow, Caffe, Darknetand DLib.• Developed strong proofs of concepts to solve real world problems like cameratampering and unusual activity detection running at over 25FPS.• Designed and maintained an end to end analytics pipeline which would facilitatecommunication between a NodeJS backend and C++ based machine learningalgorithms.

• Generated ad-hoc reports through SQL server along with the development of stored procedures.• Managed and communicated maintenance schedules of the system.