Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-3650 M: +1-331-***-****, E: firstname.lastname@example.org LinkedIn: https://www.linkedin.com/in/meet-shruti1312/ SKILLS
Masters in Computer Science, University of Southern California August 2018 - May 2020 B.Tech. in Computer Science, Uttar Pradesh Technical University, India July 2008 - June 2012 WORK EXPERIENCE
Graduate Research Assistant, University of Southern California Los Angeles, California April 2019 - Present
• Building extensive machine learning / analytical services through various approaches with F2E, Data Scientists, Data Infrastructure team
• Developed and deployed an optimized deep learning convolutional neural network (CNN) algorithm, to enable text and object detections in the Image. Trained the model on over 80,000 images to provide an accuracy of 92%.
• Developed a high quality, cost-efficient banknote processing machine. Link to product https://www.gi- de.com/en/us/currency-technology/solutions/cash-processing-solutions/banknote-processing-solutions/bps-c2-family/
• Solved critical issues in uncontrolled, non-reproducible, unbalanced data environments enabling software releases on time
• Built analytics dashboards using JSON and TCP/IP and UDP communications to remotely monitor the statistics of live banknotes processing on different machines maintaining 100% data synchronization with latency as low as 1%
• Minimized the number of software bugs to a scale factor of 30% by performing Test-Data-Driven(TDD) and Behavioral-Data- Driven(BDD) unit testing, peer code-reviews and cross-functional demos of the releases
• Liaised with a team of 6 and created a new testing process reducing the total time to software roll out by 35%
• Worked on real-time simulations of banknotes processing using TTCN framework combined with backend System Adapters, Encoders, and Decoders in python with Hardware-in-the-Loop and covered 1000+ scenarios in less than 5 months
• Followed Agile methodology and Scrum(JIRA) to meet milestones and releases, and mapped requirements to User Stories, and automated the builds and releases with Team Foundation Server(TFS) reducing 30% of man-hours PROJECTS
• Open Source availability for US Airforce: Delivered a solution prototype for US Airforce enabling to cut-short the software approval process time and money by 90%. Involved 100+ interviews with different military men to pivot in different directions and arrive at the outcome.
• Sorting Algorithm Visualizer: Visualized various types of sorting algorithms on a randomly generated array of numbers with OpenGL and C++
• Automated Car AI Agent: Built an AI agent in Python combined with Utility-Based reinforcement learning deciding its next move based on fixed rewards and penalties to arrive at its destination. Verified the correctness of calculated policy by sampling it over more than 100000 data distribution. Also, tested it against faulty mechanism by calculating probabilities