• Trained software team through presentations on software development practices including Agile, unit and integration testing, code reviews, and continuous integration• Spearheaded development of new event-based software architecture for new instruments based on Python, RabbitMQ and Qt/QML with focus on web connectivity through a JSON API• Collaborated with internal devops and Amazon AWS IoT teams in development of cloud architecture for internet-connected instruments using AWS IoT and AWS Greengrass• Mentored a junior engineer in development of a deep-learning cell classification and localization algorithm• Guided a junior engineer in design and implementation of an MVVM GUI application built using Python, Qt/QML, Websockets and ReactiveXTechnologies: Python, RabbitMQ, AWS Greengrass, AWS IoT Core, Qt/QML, Websockets, ReactiveX, Sphinx, Graphviz, PlantUML, Keras, Tensorflow

• Designed highly customized e-commerce websites using Saleor/Django, Python, JavaScript and Bootstrap while following responsive design practices• Greatly improved customer productivity by integrating an external inventory manager into Django backend• Enhanced reliability through extensive unit and end-to-end testingTechnologies: Python, Django, Saleor, JavaScript, Bootstrap, SCSS, PostgreSQL

• Researched and applied deep-learning to the computer vision problem of spoof-detection in face authentication• Achieved state-of-the-art results in face spoof-detection by applying transfer learning with deep convolutional neural networks (CNNs) using Caffe• Developed an Android camera app using Java and Snapdragon Neural Processing Engine (SNPE) to procure training data and demo the trained spoof-detection models• Automated the generation of spoof images and reduced project turn-around time by a few weeks• Created scripts using Python and OpenCV to augment and prepare training data• Implemented and optimized the deconvolution operation on an embedded Digital Signal Processor (DSP), achieving competitive performance and higher energy efficiency compared to the GPU• Practiced agile programming, scrum lifecycles, code reviews and test-driven developmentTechnologies: Python, C/C++, Caffe, Tensorflow, OpenCV, CMake, scikit/numpy, Snapdragon Neural Processing Engine, Jenkins

• Wrote Android system software in Java and C++ for Continuous Authentication• Implemented a hardware abstraction layer (HAL) to facilitate communication between non-secure Android and Qualcomm Secure Execution EnvironmentTechnologies: Android, Java, C++, GMock, GTest

• Investigated, designed and developed in C/C++ a new multicore chapter for the acclaimed Vellamo mobile benchmarking Android app• Collaborated with numerous performance teams in identifying performance issues in the next generation SOCs; these multicore benchmarks were incorporated into critical test suites used company-wide• Published a top-10 paper and presented a lecture on benchmarking insights in an internal tech conference• As a feature owner, developed an Eclipse IDE plugin using Java to visualize system traces and aid in functional and performance debugging of Android applicationsTechnologies: Android, C++, Java

• Master thesis: Improving GPU Programming Models Through Hardware Cache Coherence• First author on HPCA-19 paper "Cache Coherence for GPU Architectures". Selected for IEEE Micro Top Picks 2013 (12 top papers from Computer Architecture conferences in 2013)• Second author on MICRO-44 paper "Hardware Transactional Memory for GPU Architectures". Selected for IEEE Micro Top Picks 2011 (12 top papers from Computer Architecture conferences in 2011)

• Helped develop and maintain an open-source, cycle-accurate GPU performance simulator – GPGPU-Sim 3.x• Assisted in preparation and submission of a successful 3-year SRC grant proposal

• Designed and implemented a verification testbench environment for an Intel solid-state drive IP block using System Verilog and the Open Verification Methodology framework• Modelled and programmed block, register and coverage models along with directed and exploration test suites used to carry out the hardware verification process

• Collaborated with physicists and engineers in the development of a low magnetic field environment within a pulse tube dilution refrigerator for use in quantum computing• Designed, constructed and tested passive-compensation magnetic shields for use in the Input/Output system of the quantum computer

• Designed performance and stress tests for testing internally developed hardware and software for use in carrier grade networking equipment