I did my Masters in Computing Science (Big Data) at Simon Fraser University. I'm passionate for Data Science and competent in Deep Neural Networks. Developing custom big data solutions is my thing.
Data Scientist with more than 1 year of experience executing data-driven solutions to increase efficiency, accuracy, and utility of internal data processing. Experienced at creating data science solutions and analyzing data mining algorithms to deliver insights and implement action-oriented solutions to complex business problems. Looking to implement my academia knowledge in practical setting to manage machine learning and data-related solutions at large scale.
My previous work and projects are predominantly in Pyspark and Python. I'm skilled in PyTorch, NetworkX, and GraphFrames libraries. I'm good at writing custom ETL solutions to streamline data science operations.
I completed my under-graduation with distinction and started as an Assistant System Engineer at Tata Consultancy Services Ltd. Later, my inquisitiveness for finding patterns and crunching numbers made me shift my career to Data Science. So, I enrolled in Masters in Big Data program at Simon Fraser University to enhance my skills further. Currently, I work at Royal Bank of Canada as a Full-Stack Data Scientist.
June 2020- Present
May 2019 - December 2019
June 2016 - January 2018
September 2018 - April 2020
June 2012 - April 2016
Below are some of my recent projects which I worked on, as part of my course work at SFU. Please click on the respective projects for GitHub page and more info.
One-Pixel attack is a black-box attack aimed at fooling a Convolutional Neural Network. The attack modifies one pixel of the image and make CNN to misclassify the object. In this project, I explored how adding a variational auto-encoder can prevent such attacks.
3SA - Semantic Seacrh for Speeches in Audio, a minimally viable product that demonstrates how you can enable semantic search, specifically for speech-related audio files. Using advanced NLP techniques, 3SA allows you to search and retrieve audio files which are semantically similar to search query.
Vancouver Real Estate Analysis and Predictions. I analysed Vancouver housing market and developed a tool which leverages Machine learning to recommend and predict house prices given the area, crime rate, and other features.
This is the implementation of Markov clustering algorithm in PySpark. There is only a Python implementation of this algorithm so I am trying to scale this algorithm and make it can run on large-scale databases.
This is a basic data analysis on Yelp data. We have found some intersting insights by analysing the reviews and recommended the factors which the restraunts need to concentrate on in order to improve their business.
You can reach me through the below mentioned mails.
pavan.kosaraju@outlook.com
vkosaraj@sfu.ca