I am a graduate student, doing my MS in Intelligent Information Systems (MIIS) in School of Computer Science at Carnegie Mellon University.
I am a graduate student at Carnegie Mellon University. I am pursuing my MS in Intelligent Information Systems from the School of Computer Science.
I recently interned at BloomReach, Inc. in Mountain View as an Engineering Intern. I primarily worked on SolrCloud.
- Machine Learning (10-601)
- Speech Recognition and Understanding (11-751)
- Search Engines (11-642)
- Natural Language Processing (11-611)
- Machine Learning for Large Datasets (10-605)
- Language and Statistics (11-761)
Business Process Modeling Notation 2.0, Packaged Application Software, Programming Robotics and Sensors, Mobile Computing on Windows platform, Enterprise Application Architecture, Biomedical Informatics, Programming in C#, Object Oriented Programming, Systems Analysis and Design, PL/SQL
National Collegiate Conference 2012
National Collegiate Conference 2012
Placed first in the Boston Scientific Case Competition representing Purdue University
Won the scholarship during the Indiana Women in Computing Conference
Microsoft Imagine Cup Contest
I have designed and implemented the SolrCloud Rebalance API. This new API allows robust index and config manipulation in SolrCloud while guaranteeing zero downtime using various Scaling and Allocation strategies.It is built in a generic way to enable plugging in of new scaling mechanisms to Solr and make it contributable to the open source community.
Blog on my work: Click Here
TA for Search Engines course taught by Prof. Jamie Callan.
Click here for course website
Major Responsibilities include:
Writing Python and Perl scripts/code for auto-grading coding assignments
Generating tests cases for homework assignments
Managed the design and analysis for new strategic initiatives
Designed and implemented data warehouse for the enterprise repository
Taught technical courses in Agile, SQL, Object Oriented Programming concepts
Managing over 25 Lab Consultants and their day-to-day activities
Support all the administrator duties including scheduling, approvals and shift reports
Monitor technology laboratories
Developed a dynamic form builder with drag and drop functionality as a plugin to various different types of industry applications
Developed 10 features for the July 27th Production release for the Access Request Management portal
Completed the first draft of the data warehouse architecture design.
Worked for the Associate Department Head as a Web Developer
Managed all the accounts and budgeting
Application Development for the website
Worked on the NSF-ITEST funded SPIRIT (Surprising Possibilities Imagined and Realized through Information Technology) project
Developed features for various applications
On-boarded 69 internal bank applications on Access Request Management portal
Implemented UML 2.0 in the system documentation
Developed an application for the New York museum special exhibits’ Virtual Tour of Artifacts
Worked for the Dean of College of Technology
Developed surveys using Qualtrics software
Gained valuable communication skills due to constant interaction with high officials and non-technology related people
We created a Question - Answering system, that could generate coherent, competitive and grammatical questions, given a Wikipedia article. Additionally, the system could answer questions about a given article. We used several different NLP techniques for this implementation, including dependencies like NLTK, Stanford Parser, MITIE etc. I was responsible for the preprocessing module and the answering module. Our team won the best answering system at the end of the semester.
Team Members: Sai Sumanth Miryala, Bhawna Juneja, Pradeep Karuturi
Youtube URL: https://www.youtube.com/watch?v=0Sk2xMqLikw
Using topical and contextual information to create personalized email sorting system as a part of the Yahoo!-CMU collaboration. I used clustering techniques to generate a model per user that separates different types of emails and sorts them based on learned user preference.
Using speech data to detect driver distraction. Created and maintaining a driving simulation system (using Java and OpenDS) for data gathering. Controlling the human interactive system to vary cognitive load to collect speech data.
Project URL: http://utc.ices.cmu.edu/utc/projectitem.asp?ID=97
Using the KALDI speech recognition toolkit
Training grapheme to phoneme converters to generate pronunciations of twitter data
Using clustering techniques (Percy Brown's Brown Clustering) to group similar sounding words together
Improved lexicon pronunciation dictionaries.
Team Members: Jeremie Kim, Mrigesh Kalvai, Vivek Pai
Created Text Search Engine using a pre-indexed corpus consisting of 10% of all Wikipedia webpages (Lucene), as part of the Search Engines and Web Mining courset
Created parsers which would be able to handle structured queries consisting of operators like 'AND', 'OR', 'NEAR', 'WEIGHT', 'WINDOW' as well as handle Bag of Words(BoW) queries
Implemented Retrieval models such as Unranked Boolean, Ranked Boolean, Okapi BM25
Implemented statistical models like Indri with pseudo-relevance feedback
Built and evaluated a robust classifier learner for object detection using the CIFAR-10 image dataset using Machine Learning techniques
Implemented an "Ensemble of Ensembles" classifier for training the model
Used feature extraction techniques (HOG Feature Extraction and Principle Component Analysis) for data preprocessing. The team ranked 6th (out of 123 teams) in Kaggle competition for the same.
Team Members: Lara Martin
Implemented Logistic Regression model whilst using SVMlight for Learning To Rank (LETOR). Evaluated models employing rank-based evaluation metrics - Normalized Discounted Cumulative Gain (NDCG), Mean Average Precision (MAP) and Precision@N (P@N). Implemented regularized logistic regression (RLR) with gradient descent, and ran experiments on adapted RLR and SVM for LETOR, on a dataset from Microsoft Research Asia.
Developed a desktop application to store and manage patient records, including medical history, drug dosage thresholds, insurance information and other relevant data. Also, created an access management system user interface.
Developed a mobile app for Windows phones (WP7) primarily for asthma patients. A pacifier appcessory is attached to the audio jack and detects the intensity of air blown into it. The game uses the intensity to drive away the obstacles from the animated moving path. Performed research and discussions with 30 patients and 2 doctors to validate the design.
Created a web application for Windows phones (WP7) using open source speech-to-text translation library. The app converted text into animated sign language, helping the speech and hearing impaired individuals to communicate. The app framed was designed to be extended as a plug-in to video conferencing tools such as Skype and Google Hangout.