Jan 2019 – May 2019 · 4 mos

• Teaching Assistant for Computer Systems Course CSCI 2400

Sep 2018 – May 2019 · 8 mos

• Working on Elastic OS(EOS). EOS supports scaling of applications in a cluster without needing to rewrite them. This requires modifying the linux kernel.

Aug 2018 – Dec 2018 · 4 mos

• Teaching Assistant for Data Structure Course CSCI 2270

Jun 2018 – Aug 2018 · 2 mos · Greater San Diego Area

• Part of the Platform tools and Engineering Group. Working on Docker, Kubernetes, Jenkins.

Jan 2018 – Apr 2018 · 3 mos

• Student Assistant for Discrete Structure Course CSCI 2824

Aug 2017 – May 2019 · 1 yr 9 mos

• Master's Student in Computer Science.
• Subjects completed:
• 1. High Performance Scientific Computing
• 2. Natural Language Processing
• 3. Distributed Systems
• 4. Internet of Things and Wireless Systems
• 5. Machine Learning
• 6. Data Center Scale Computing
• 7. Advanced Algorithms
• CGPA: 3.97 / 4.0
• HIGH PERFORMANCE COMPUTING
• I took High Performance Computing (Parallel Computing) as a part of my coursework. Through this course, I have learnt several parallel models and networks and used these parallel techniques while implementing different algorithms. I designed and developed iterative solver for Poissons Equation employing the Parallel Mesh model and using OpenMPI for implementation.
• The research project I worked on was a parameter identification problem in the field of Pharmacokinetics. The problem definition was to find the best set of parameters of an ordinary differential equation where each parameter represents a certain weight for a particular behaviour. The differential equation would be a function of a variable and time. The problem involves estimating as many as 50-60 parameters. Running the best sequential algorithm would result in a huge running time. In order to make this perform better, I converted the sequential algorithm into a form on which parallelism could be applied easily. This was approached by employing a discretization method (Forward Euler/Backward Euler) and parallelising this discretization process for different step sizes for a given parameter set. The result would be a set of parameters which would stabilize the differential equation whereas all the parameters for which there is an instability are ignored.
• NATURAL LANGUAGE PROCESSING
• Natural Language Processing course has equipped me with fundamentals of language models, language generation. I have worked on projects like Sentiment Analyzer, True/Fake Statement Classification, Part of Speech Tagging and IOB Tagging.

Jul 2015 – Jul 2017 · 2 yrs · Bangalore, India

• Part of the System Software Team at Amagi.
• REAL TIME TEXT FEATURE
• Responsible for the development of a real time text feature on the graphics framework (handles display of graphics like logos, astons) for all the different types of custom hardware at Amagi. This feature allowed television broadcasters to dynamically display text over graphics such as tickers and counters, with minimal delay, instead of sending thousands of images with embedded text to the hardware, thereby reducing the bandwidth and memory usage.
• PLAYLIST MANAGER
• Designed and Developed Playlist Manager (PLM), a software which schedules
• the contents to be played out on television from a playlist given by the broadcaster with the support for Live Content Scheduling. Development was done in C and Python.

Jan 2015 – Jun 2015 · 5 mos

• LOG COLLECTOR AND ANALYSER
• Developed Log Collector and Analyser (LOGAN), a tool built to monitor and analyse real time
• logs of an application. I was involved in building this system from the ground up, extensively using design patterns in the process. An MVC architecture was employed to collect, store, and analyse logs. The process of seamlessly collecting logs from different application clients and transmitting these logs to a server required me to design a communication system on the client side as well as an efficient database design on the server side.
• EMPLOYEE MANAGEMENT TOOL
• Developed an in-house tool aimed at monitoring employees’ activities on different platforms like Jira, Github, and Google Apps in the form of a dynamic dashboard. Development was done using Django, AngularJS.