Projects


Prevention of DNS Cache Poisoning Attack | Mentor: Dr. Vishal Gupta

August 2017 - December 2017

Worked under the guidance of Dr. Vishal Gupta in exploring various methods to prevent DNS Cache Poisoning Attacks and figuring out an implementation in Software Defined Network(SDN) environment.

We explored various methods of cache poisoning attacks and their prevention methods and tried to figure a solution to prevent them in SDN environment. Although DNSSEC is a fullproof solution to cache-poisoning but its complete adaption is yet to take some time.

We developed and implemented a solution based on idea described in a paper Delay Fast Packets. The solution was written in FloodLight a SDN Controller and tested on a testbed written in MiniNet.


Extractive Text Summarization | Mentor: Dr. Yashwardhan Sharma

August 2017 - December 2017

Worked under the guidance of Dr. Yashwardhan Sharma in WiSoC Lab to conduct studies in the field of automatic text summarization especially techniques used in the domain of Neural Networks.

We developed a prototype system based on CNNs for the task of summarizing texts of Indian Languages as a part of this project and also a para-phrase detection system for training data generation due to lack of enough corpus on Indian Languages like Hindi,Bengali,Marathi and Telugu.

I also developed a small tool for scraping and transliterating various tweets and prodct reviews for an on- going sentiment analysis project in the lab.


Backend Platform | Legistify

Feb 2015 - December 2015

Worked with Legistify which is a startup from BITS-Pilani as one of the initial members of the team to setup the MVP for the legal platform.

We developed the MVP and core feature of automated legal document generation to develop a one stop, hassle-free place for all major legal documents and agreement types generated and delivered by just providing the basic information needed.

Also developed the first version of a Q&A style platform which helped people post their common legal queries and get answers from registered lawyers on the platform.


Mini Projects

  • Codeforces Rating Predictor : Linear Regression based predictor for Codeforces Rating based on user history.

  • Glyphicon Character Recogniser : LeNet style CNN based architecture for glyphical character recognition.

  • Parallel Branch and bound : implemented branch and bound for Travelling Salesman Problem and Knapsack Problem.

  • Blockhain-TS : A naive implementation in Typescript to understand the basic of BlockChain which powers most crypto-currencies and crypto contracts platforms.

  • Developed a compiler from scratch for an dummy language called ERPLAG.

  • Contributed code analyser module to phpsa, a open source project.