I am a Software Engineer at Manulife. Before this, I was a Computer Science grad student at University of Waterloo. Earlier I was a Software Engineer at Johnson Controls. Even earlier, I was an under graduate at VIT, Vellore, where I majored in Computer Science and Engineering. Interests: Algorithms and Machine Learning. I have done several independent projects in the area of web-development, big-data and machine learning (explore)

Research

Publication in Applied Soft Computing (Worldwide Rank 9 in AI). See full Google Scholar profile here.

Industry Experience

I work with Advanced Analytics team at Manulife. A brief summary of my work:

• I am helping to reduce the employee attrition within the company. I built an application from scratch for analysing organization wide network of emails using Pyspark. Detected communities within that network. Found influential employees within those communities who drive employee engagement. This application reduced the communication lag between employees by ~30%. It also helps in improving HR operations within the company to reduce employee churn. Besides this, I also built a model for predicting employee churn using features like distance from home, textual surveys, sentiment scores, LDA topics, etc.
• Besides this, I worked to optimize the utilization of workspaces and meeting rooms within Manulife's offices. This was done by utilizing the data of sensors installed at employees' workspaces and data from employees' outlook calendar.
• I built a pdf scraper to scrape and analyse data related to specific tags from a repository of pdfs.

As a part of big-data team (Connected Services) in Johnson Controls, I developed several map-reduce jobs (for problems like computing device run hours, finding hourly energy consumption) to process terabytes of data emitted from chillers (giant machines which regulate the heating, ventilation and air-conditioning in buildings). Also, I collaborated with the team at Milwaukee to create a subset of the back-end of company's flagship product called Metasys. In 2016, I won the runner's up award at Johnson Control's technical fest for prototyping an idea which aimed at predicting faults in chiller operation, this consequently led to filing a US patent application (Gallery)

Skills

• Langauges: Python, Java, Latex, Javascript, SQL, HTML
• Frameworks: Hadoop, Spark, Gensim, Tensorflow, Django REST
• Apis: NumPy, Scikit-learn, Pandas
• Databases: MySQL, Neo4J, Hbase, Hive
• Operating Systems: MAC, Linux, Windows

Relevant Courses

Graduate:

• Software Engineering for Ultra Large Repositories
• Deep Learning

Undergraduate:

• Data Structures and Algorithms
• Linear Algebra
• Applied Probability, Statistics and Reliability
• Multivariable Calculus and Differential Equations
• Differential and Difference Equations
• Data Warehousing and Data Mining
• Operating Systems
• Algorithm Design and Analysis
• Object Oriented Paradigm
• Database Systems
• Software Engineering