Hi, I'm Shubham. I am a Graduate student at Stony Brook University.
I am very passionate about programming and engineering as a whole. I primarily work on the backend and I am very fond of complex systems that require critical thinking skills and logic. Though, I am flexible and have worked in the frontend as well!
I recently developed an interest in Machine Learning and Data Science. I learn new stuff everyday!!
I am currently interning at Remote Roofing as a Machine Learning Intern. I previously worked as a Software Engineer at Robert Bosch Business and Engineering Solutions.
Areas of interests - Machine Learning, Data Science, Theory of Databases, Big Data Analytics.
Personal Interests - Soccer Buff( Glory Glory Man United!), Meeting and Interacting with new people, Exploring new places.
Write modern, performant, maintainable code for a diverse array of client and internal projects.
Communicate with multi-disciplinary teams of engineers, designers, and clients quite freqeuntly.
Fabricating web applications in Vanilla Javascript, Jquery, Ajax, HTML in use by 30 employees, resulting in
10+ new features, reduction of 50% in save/load time and 25% operation time.
Implementation of Deep Learning using Mask R-CNN, MS COCO, ResNet101 to remotely inspect and assess roof damage. Identified the roof damage with an accuracy of 96.58%.
Stony Brook University
August 2019 – December 2019
Teaching Assistant
Orchestrated the TA duties of 250 under-graduate students under Prof. Michael Tashbook.
Robert Bosch Engineering and Business Solutions
October 2017 – June 2019
Software Engineer
Supervised a team of 4 in Germany as Functional Safety Developer, to achieve 80% reduction in testing time
by developing autonomous testing software for fault management of CAN Frames.
Proposed and implemented scalable solutions to issues identified within the software services and applications responsible for Vehicle Safety.
Executed responsibilities on various requirements like Tester Diagnostic Services, CAN frames, Cruise
Control functionality, Driver Demand Requests. Remodelled the safety software for complex systems like
Engine Control, Remote Parking Assistant, Handling Transmission and Vehicle Stability interventions; increased the
throughput by an average of 20% and reduced customer complaints by 50%.
Education
Stony Brook University
August 2019 - PresentMaster of Science in Computer Science
Courses Taken: Machine Learning, Data Science, Artificial Intelligence, Big Data Analytics, Theory of Databases, Analysis of Algorithm.
Grade: 3.7/4/0
KLE Technological University
June 2013 - July 2017Bachelor of Science in Electronics and Communication
Co-founder of Forum for Impacting Social Transformation(FIST) - A social group which aims to make a positive impact on the unpriviledged people. This group aimed to create awareness about the social problems and found a way to use technology and youth power to alleviate the social problems.
Designed and implemented an information retrieval and classification system for sentiment analysis on Twitter. Crawled tweets on user’s timeline from Twitter API, and extract JSON responses using Requests module.
From the script generated, the Tweets were directly loaded in an AWS Kinesis Firehose delivery stream. I used Lambda function for the transformation. The sentiment information was obtained using Amazon Comprehend and finally the Tweet and its sentiment data was stored in an Elasticsearch domain. Through this, various real time informations were analyzed.
Created a data pipeline which extracts real time tweets using Twitter Stream API. Used AWS Kinesis FireHose, S3 bucket and Amazon Elastic MapReduce(EMR) to stream, store and analyse the Tweet Data. Utilized PySpark jobs to load tweets from S3.
Performed an in-depth analysis on the census data of Florida and New York state in order to obtain insights on the impact of political party affiliation of couples on their marriages and its correlation to the divorces using Pandas, Numpy and Sklearn libraries.
Virtual Camera Application
Worked on a Containerised Environment to develop an API which processes the Image Sensor details to store the image characteristics in the Database using Python,Flask and Docker
Map-Reduce Library
Implemented a Hadoop-like Map-Reduce facility,with master and worker nodes for map reduce operations over large datasets,with a distributed file system,and fault tolerance to address datanode failures
Effectively built a Recommended System using Item-Item Collaborative filtering method and recommendation using minhashing method (Similarity Method) using Apache Spark and HDFS.
