I am an experienced Java developer who is good at writing efficient algortithms. I have coded for a variety of projects, both in the industry and academics in a range of different languages.
I like to believe I have good communication skills. I have been a scrum Master, interviewed folks, good storeyteller and even occasionaly writes answers on Quora for fun.
I am really good at designing the use cases and workflow of an application and almost always delivers more than promised. I like to take ownership and do not rest until I make it the best out there.
An application layer protocol for client-server model to communicate between server and multiple clients in private and public mode using socket programming in python.
Created a website using the Django framework to deploy a Python app. Ensured it stays up for the duration of a term while surviving real-world issues. Used tools like supervisord, kibana and elastic search for deployment and monitoring. The website was load balanced so that it can withstand sudden traffic surge by automatically deploying additional cloud computing resources.
I created a Java Application that utilizes MongoDB Compass to process a very large data set of transportation and roadways department. The application uses Mongo queries to query data in JSON format and answer questions determined by the project description. The project also utilizes indexing to speed up frequent queries.
I created the Official website of the Indian Student Association at Portland State University. Website
A smart supply chain IoT solution to provide real time information to Warehouse managers. It provides automatic scheduling of trucks, automatic parking allocation and real time status of inventory loading/unloading. The solution harvests the IoT capabilities of SAP Hana Cloud. This project won the first prize in SAP India and 5th rank in SAP Global IoT challenge.
Filter and sort all University Transportation Center (UTC) funded research from trid.trb.org that relates to bicycle and pedestrian research and display it in an easy to use format on a single site. This application betters the filter functionality for existing research by sub-categories, enabling the end-users to quickly access and find the information they need.
An application which allows authors and users to review books and participate in discussion forums. Users can create a personalised experience by creating their own profile. The application utilizes various technologies like Bootstrap, JQuery, Ajax and Google API's.
This design project gave me a hands-on experience with a variety of techniques for HCI design. I engaged in substantive individual as well as group work, where I successfully conducted interviews, obervations and market research to identify problems and further prototyping and storyboarding the final design solutions. More Description
Yoda Chatbot is a machine learning application written in python that mimics written human speech for the purposes of simulating a conversation or interaction with a real person, in a manner that Yoda would. It uses recurrent sequence-to-sequence model. GitHub
Leading the national non-motorized count data archive, BikePed Portal provides a centralized standard non-motorized count database for public agencies, researchers, educators, and other curious members of the public. A few of the upcoming features include a user restricted service for data owners to QA/QC their data, download raw data, and export data to the Traffic Management TMAS.
The project utilizes twitter data on trending economic topics on Twitter and gauges the overall user sentiment. Then by performing sentiment Analysis in R language, we determine average user sentiments on a number of days. The project then uses a regression technique to determine a correlation between the twitter sentiment and the value of INR (Indian National Rupee) on the following days. The scope for this project was the twitter data from the demonetization phase in India.
In recent years Runtime Diagnosability has emerged as a hot Operating Systems research topic. This includes such sub-topics as performance measurement, tracing, profiling, debugging, bug localization, error detection, bottleneck analysis, white-box testing, failure reproduction, fault detection, automatic repair, etc. Researchers have published several papers about it in high-quality conferences such as OSDI, SOSP, SuperComputing, PPoPP, ASPLOS, and others. We identified the top 20 most important high-quality research papers in the area of Runtime Diagnosability using a range of metrics like h-index, acceptance rate, etc.