Generic placeholder image


Computer Science Undergraduate

IIT Bombay

I am a final year student studying Computer Science at IIT Bombay.

I have a keen interest in Artificial Intelligence and Machine Learning. I also like to analyze! Studying data and extracting insightful information is what I love to do. Also, as a side-interest, I have great enthusiasm in gaming and would look forward to making AI for many games.

Other than my professional interests, I am a big-time sports follower. I love to play football and basketball. I am a Chelsea fan. In basketball, I do follow NBA and my favourite team is Cleveland Cavaliers. Other than that, I like to read books. And not only reading, but I like to write poems. (You can find plenty of them at my blog).

All in all, I have feet in various frontiers and am trying to balance all of them. That's all about me.

P.S. - You can contact me here.


Following ones are open and research oriented projects and interns

my projects

Speech Recognition Systems for Code Switched Speech (in collaboration with Microsoft Research)

Guide: Prof. Preethi Jyothi

Code-switching is a practice of using two or more languages in a single utterance. Due to lack of significant amounts of clean and labelled data, a robust speech recognition system can't be built directly. We explore techniques of using monolingual language models of the individual languages and combine them in a way to build a combined language model which can then be used to build a robust speech recognition system. Future work includes exploring usage of linguistic rules and encoding them to be able to generalize better.

my projects

Opponent Modelling in Scrabble

Guide: Prof. Shivaram Kalyanakrishnan

We look to design an inference agent to model the opponent's rack. Using information about the opponent's rack, the agent can play aggressively or defensively. We use simple Bayesian Inference based model to devise a probability distribution over possible racks. We impose further constraints by introducing information about previous opponent's racks. We use this information in forward Monte Carlo simulations to find better moves.

my projects

Defect Detection from Product Reviews

Guide: Prof. Pushpak Bhattacharyya (in collaboration with Accenture)

We aim to find the specific sentences from product reviews which directly delineate the defect of the product from the review. We first build a product ontology which captures the defect-related phrases and words. For any candidate test sentence, we use a pattern matching approach to find the similarity with the ontology. We use POS tags and word embeddings for better generalization.

my projects

Interactive Virtual Chat Bot

Philips Research, India

The aim was to develop a chat bot for medical applications. I worked on several parts in the same, one of the important ones included the problem of lip-syncing. This included studying various co-articulation models to obtain the best mapping of visemes with an additional constraint of ensuring smooth transitions. Apart, I also worked on building the framework for scene recording and generation dynamically.

my projects

Improving Language Models using Cross-Scripted Text

Microsoft Research, India

Cross-scripted text is the transliterated version of the native language text in some other language. These can act as secondary sources of data, especially in contexts where native language text is limited. We aim to exploit this data source using a robust transliterator and using the transliterated text to improve the language models for the native language.


My Recent Couse and other Projects. Click on them to know more

my projects

Deep RL Agent for Bomberman

Foundations of Intelligent and Learning Agents | Prof. Shivaram Kalyankrishnan | Python, TensorFlow

We explored various learning methods like SARSA, Q-learning. We used a simple neural network for Q-function approximation. We also focused on improving the quality and time of convergence of Q-values for which we used human-based features. We also used the idea of Curriculum Learning to teach the agent complex tasks in a simpler and faster way.

my projects

Audio based Sentiment Analyzer

Automatic Speech Recognition | Prof. Preethi Jyothi | Python, OpenEAR

We used OpenEAR to extract features from audio files and further used feature selection to reduce the dimensionality of the features. We trained classification models using these features for sentiment prediction. We also explored an ensemble model combining the text based sentiment analyzer for which we used word embeddings. We tested these models on various data sets and proved improvements over the baseline.

my projects

Intelligent Agent for Pacman

Artificial Intelligence | Prof. Shivaram Kalyanakrishnan | Python

We used a basic implementation of the game of Pacman. We implemented and compared various simple heurisitics like search, reflex agent, Minimax algorithm, Alpha-Beta Pruning and Expectimax algorithms. In order to maximize the performance of the agent, we used heurisitics to enhance the evaluation function. We explored another case where the ghost position is unknown. We implemented and tested Particle Filters and Dynamic Bayes Net to infer their position.

my projects

Social Reader's Platform

Databases | Prof. S. Sudarshan | Java

Inspired by goodreads, we aim to build a social platform for readers wherein you could share, recommend, find books. It will be a web-applet wherein the front end will be Javascript, Ajax, etc and the back end databasing would be done in postgres.

my projects

Diep 2.0

Self Project | Unity

My first major gaming project. It's an attempt to replicate diep.io. Rather, we plan to add additional features and improve on the latency part. It's an ongoing project and plan to launch it for the institute soon enough!

my projects

Malicious URL Detector

Machine Learning | Prof. Ganesh Ramakrishnan | Python

Idea is to detect whether a given url is malicious or not. To start with, we used the extracted lexical and host-based features of the url and trained various classification models on them. We ended up with a accuracy rate of > 95%, with testing done on more than 20000 urls.

my projects

Traffic Management System

Digital Logic Design | Prof. Supratik Chakraborthy | VHDL

We planned to take a real time traffic management example as of like the three signals outside IITB. Work included the synhronizaton and timing of the lights in order to avoid ever-increasing waiting time and smooth flow of traffic.

my projects

Socket Programming

Computer Networks | Prof. Kameswari Chebrolu | C++

After learnt the basics, it was time to implement them. The setup consisted of a single server having many worker connected to it and multiple clients trying to connect at the same time. Client requests to crack a password and the workers run brute-force algorithms to crack 'em down.

my projects

Branch Change Portal

Software Systems | Prof. Sharat Chandran | Django

IIT has allowed changing of branches after the first year. So, the project was to build an online portal for the students to register and fill in their preferences. Then, we also implemented a Gale-Shapley kind of algorithm to allot the students their respective branches according to the set preferences.

my projects

Rube Goldberg's Model

Software Systems | Prof. Sharat Chandran | Box2D, C++

This was a simple simulation of the Rube Goldberg's model. The concoction served a glass of water. The project was aimed at getting us sufficient practice to heavy-level coding.
You could check it out here

my projects

Ultimate Tic-Tac-Toe

Introduction to C++ | Prof. Deepak Phatak, Prof. Supratik Chakraborthy | C++

After learning the basics of C++, this was the first major project taken. We had built a two-player model initially where in you could play the game and the algorithm underneath detects win/loss. Later we added a 'v/s CPU' mode, wherein you would play against a normal AI.

my projects

Cricketing Model

23rd Yard | Techfest, Self Project | MS Excel

This was done as an entry to a data-driven cricketing competition at Techfest. We secured the 1st place in this competition. We had formulated a probabilistic cricketing model to predict the winner of the ICC World Cup ’15 using the data of all teams and players of the past 6 years. We also made a dynamic analytical model which would predict the rankings of players and teams.


Have a look at my resume -

Seems you don't have the required pluggin :( click here to download the PDF file.


I will be delighted to get in touch.