World!!

Hello, I'm Sudhanshu Bahety. I'm a 1st year Graduate student at Department of Computer Science and Engineering, UC San Diego . I had completed in Undergraduation from Department of Computer Science and Engineering, IIT Kharagpur.

I've recently worked at Headout as a Full Stack Engineer Intern during Summer, 2016. Before that I was a research intern at Adobe Systems during Summer, 2015.

Eat. Code. Enjoy.

Some of my past experiences

I was a Full Stack Enginner Intern at Headout. I started off my internship by working in iOS Application by designing and writing logic for Accounts tab in the application. Account tab includes Login, Logout option via Facebook, Wallet History of the user, Earning Credits and other options for the user. Apart from accounts tab, I learnt how deep links and deferred deep links work in an application, which was used for referral credit system. I got to learn about the MVC Architecture which was a plus point from the internship. There were also alot of minor changes in which I made my contribution such as integrating Algolia search, designing and writing logic for explore feed when the user is logged in or not logged in, improving city search.

I got my hands dirty by working in the mobile web application and backend side of things at Headout. Adding functionality such as adding a YouTube video for a product, improving schedule displayed for a product were something that I implemented. Improving websites metadata and sitemap for better indexing and Search Engine Optimisation, tracking user events for web push notifications were also a part of my project. A small project in my plate was to crop and upload multiple images on the go when content of a product was being added/modified.

Working as an intern in a startup like Headout was really one of the best experience you could ever have. You have complete freedom over what you do and how you do it. You can convey your own idea(which is highly appreciated) and how it would help in overall growth of the company. There is a flat hierarchy present throughout the company and you can go and talk to anyone anytime, be it CTO, CEO or COO. It will be common to see CTO will working late at night till 4 and sometimes the entire night. Everyone in the company are really passionate about their work and their contribution. They definately work hard but party much much harder. I was lucky to be a part of 7 day workation(Work + Vacation) with Headout at Goa \m/ (during my first week at work).

Tech team at Headout was one of the best. We not only focussed on what would help the product but we also had an excitement to learn new technologies. There used to be daily standups, WTFs, session on wide array of topics such as Android, RxJava, Spring MVC, Optimisations on web development and a lot more is still going on! I had joined the company with an expectation to do something out of the box, something exiciting and to contribute significantly to the product, and Headout gave me more than I expected.


At Adobe Systems, I was a part of Adobe Analytics Team. We defined our own problem and then went about solving it. Our team worked on Context Aware Media Content Analysis, in which we predicted the context in which a media story would be read. We defined the Context of a user as one of the following activity - sitting, lying down, walking, transit, running.

During my internship, I also developed an Android Application - ProReader to test our hypothesis. The hypothesis that we thought was that users tend to read different media articles based on its Context. The analysis was done mostly in R.

The experience gave me an exposure to applied research that goes on in big industries. How should you go about defining your problem statement, solution approach, gathering data and verifying your results are most important questions that you need to answer. At Adobe, I was fortunate to develop an acumen for applied research.




At Indian Statistical Institute, Kolkata, I attended lectures on basics of cryptography, and also learnt about basic practical crypto tools.

I also worked on a small project under Dr. Sushmita Ruj. The project was was to implement a paper - Key Management Scheme for Distributed Sensor Networks. Sensor Network can be considered as an undirected graph with source and destination vertices and we have to send information through this network.

We analysed how the network parameters like diameter, average degree changed when a node is compromised by an adversary. The implementation was in C++ using networkX Library. The internship at ISI gave me an initial exposure to the field of cryptography and its applications in cyber security.


Research is not boring!!


Early Detection of YouTube Spam Videos

I'm working with Dr. Pawan Goyal and Dr. Saurabh Bagchi on detecting the spam videos that are present in YouTube.

We have crawled data from YouTube and trying to analyse various video, network and comment features to classify a video as spam or not.

We used YouTube DATA API v3 to crawl the data and using python along with sklearn toolkit to build our classification model.


Behavior of Alcoholic Drinkers in Twitter

We studied and analysed the behaviour pattern of users in Twitter. From the sentiments of the tweets and keywords used in the tweets, we found a difference in tweeting pattern of normal user v/s drunk user.

The different tweeting pattern were seen based on kind of swear words, food words, health words that drunk user used compared to normal user. Moreover, we also saw a temporal pattern in the use of tweets by drunk users compared to normal users.

The final aim of the project was to tag user as fake or genuine based on their tweeting pattern. Currently, we are doing further analysis of the tweets and trying to improve our result.


Estimation & Prediction of writing style of users on Quora

In this project, we studied and analyzed the evolution of a user’s answering style over a period of time on Quora and developing a model to predict it.

We had modelled the writing pattern of a user in terms of Humor, Readability, Inappropriateness, Formalness, and Detailing. We had dataset which had over 1 million answers, and around 2000 users. From this dataset, we selected frequent users those who are regular on quora and did analysis on these users.

After analyzing writing patterns of users, we built a model to predict the answering pattern of the user in the future. All the documentaion and codes of the project are available over here.


I like to develop utility softwares


ProReader

Developed an Android Application during my Summer Internship, 2015 at Adobe Systems. It is a reading application for users who love to read news from different sources. It was built as a part of research project to collect user data.

View in Play Store »


Locus

This is an Android Application to track a BLE beacon and sound an alarm if the beacon goes out of bluetooth range. The beacon which works with this app is a Nordic BLE.

This Project started off as as a Self Project idea. The idea was to track the personal belongings of a user. The beacon is a coin size device which is attached to personal belongings such as car keys, wallets, and then it can be tracked using the cell phone.

If at any point of time the beacon goes out of range, an alarm is beeped as a reminder to the user like "Hey! You've left you're wallet! Go and pick it up".

Patent titled “A system and method for range independent tamper proof tracking” is pending for this project.
Patent Number :295/KOL/2015


We are a team of four. Other members are Manav Kedia, Kumar Abhishek, Vatsalya Chauhan.

View source code »


IPL Auctioning System

Our team had built an IPL System as a part of our DBMS Project.

The system can be used during the auctioning of Indian Premiere League. It contains the data of all the teams that are a part of IPL and data of individual players as well.

The data was crawled from IPL Wikipedia & IPL official website. The system was desinged to answer various queries pertaining to players or teams.

View source code »

Black Jack

Implemented a Web application game of Black Jack
during a hackathon organised by Mozilla.

View source code »

Online Sales Portal

A software to buy/sell products online. The idea is similar to eBay or OLX.

View source code »



Find me on Github for more open source projects I have created and contributed to.


I love ...

Developing Software

I love solving problems that most of us face in daily activity. With Software you can solve a problem as small as event scheduler to solving problems like buying products online.

Travelling

I am very adventerous. I love exploring different culture and meeting new people. My most memorable experience has always with my friends when we went to Pelling and Bangkok.

Music

I love listening to music as well as playing music. I play keyboard in part time. You can find a short compilation of Raabta over here. I'm playing the keyboard, and Sunny is singing.


Academics


Undergraduate
2012-16


Indian Institute of Technology Kharagpur, India

B.Tech. in Computer Science and Engineering

CGPA: 9.50 / 10
SGPA: 9.64 / 10


High School
2011


Calcutta Boys' School

Indian School Certificate

91.5%


Middle School
2009


Calcutta Boys' School

Indian Certificate of Secondary Education

90.2%

At IIT Kharagpur, I've completed the following set of courses.

Computer Science
  • Artificial Intelligence
  • Machine Learning
  • Natural Language Processing
  • Information Retrieval
  • Image Processing & Computer Vision
  • Complex Networks
  • Social Network Science
  • Operating Systems
  • Computer Networks
  • Distributed Systems
  • Database Management Systems
  • Programming & Data Structures
  • Algorithms
  • Approximation Algorithms & Design
  • Software Engineering
  • Compilers
  • Theory of Computation
  • Formal Language and Automata Theory
  • Discrete Structures
  • Switching Circuits and Logic Design
  • Computer Organisation and Architecture
Mathematics
  • Matrix Algebra
  • Linear Algebra
  • Mathematics-I
  • Mathematics-II
  • Probability & Statistics
Electronics & Electrical
  • Introduction to Electronics w/ lab
  • Signals and Networks w/ lab
  • Electrical Technology w/ lab
Other
  • Environmental Science
  • Science of Living Systems
  • Physics w/ lab
  • Mechanics
  • Chemistry w/ lab
  • Engineering Drawing and Computer Graphics
  • English for Communication
  • Manufacturing Processes Lab
  • Sociology of Science
  • Sociology of Economic Development