Sai Teja Duthuluri is a computer science graduate with a M.S. from Georgia Tech and B.Tech from National Institute of Technology Karnataka. He has work experience as a research assistant at Georgia Tech developing machine learning algorithms and as a software developer intern at Pindrop and Oracle India. His technical skills include Python, C/C++, Ruby, and machine learning techniques like PCA, LDA, and particle/Kalman filters. He has also completed projects in medical records, music recommendation, visual SLAM, and music genre classification.
1. Sai Teja Duthuluri
dsaiteja1908@gmail.com
(404) 953 – 9489
EDUCATION
M.S. in Computer Science Aug 2015 – Dec 2016
Georgia Institute of Technology CGPA: 3.87/4.0
B.Tech. in Computer Engineering Jul 2008 – May 2012
National Institute of Technology Karnataka, India CGPA: 8.23/10.0
TECHNICAL SKILLS
Advanced: Python, C, C++, Ruby
Proficient: Perl, Shell scripting, PHP
Technologies: REST, Django, Selenium, MySQL, Cucumber, RSpec
MACHINE LEARNING SKILLS
Experienced in CV techniques like Harris corner detection, SIFT, HoG, RANSAC, Bag of Words, Viola-Jones face detection
Experienced in applying PCA, LDA, NB-Classifier, Ensemble learning, clustering techniques towards image classification
Implemented supervised learning and unsupervised learning methods, tested on public datasets
Implemented particle filters, Kalman filters for robot localization and online learning algorithms for real-time data classification
WORK EXPERIENCE
Graduate Research Assistant Aug 2016 – Dec 2016
Georgia Institute of Technology
Maintain and develop new features for an open-source Lab Information System, developed in PHP with MySQL and Apache
backend
Currently deployed in multiple African countries
Software Developer Intern May 2016 – Jul 2016
Pindrop
Developed a Ruby-based framework that enabled UI and back-end test automation
Thoroughly followed Agile philosophy, Test-Driven Development, Behavior-Driven Development principles
Graduate Research Assistant Jan 2016 – May 2016
Georgia Institute of Technology
Isolated SSVEP EEG brain signals using in-ear electrodes
Developed end-to-end pipeline for signal acquisition, processing, peak detection and classification in Python
Software Developer II Jun 2012 – Jun 2015
Oracle India
Developed test framework and tools for OVM for Sparc, a virtualization product for Oracle Sparc Enterprise Servers
Developed test framework for power management features of the Oracle Sparc Enterprise Servers in C and Perl
Software Engineering Intern May 2011 – Jul 2011
Cisco India
Prototyped a patch for SystemTap which enables placing hardware watchpoints for user-space variables, using the perf
subsystem in Linux
PROJECTS
ConCat – Electronic Patient Record System PHP, HTML
Developed a PhoneGap application for android and a website for tracking medical records of congenital cataract patients
Currently deployed in African hospitals
Developed REST APIs and deployed on AWS
Music Recommendation System Python, PHP
Aggregated low-level song features into song signature and compared songs using Earth Mover’s Distance (EMD)
Developed the data processing, feature extraction and similarity metric pipeline for the system
Loop Detection in Visual SLAM Python, Caffe
Fine-tuned existing models through a deep learning Siamese pipeline and custom robot image data
Used the tuned model to build a binary classifier to detect loops in robot path using only the images captured (~90% accuracy)
Music Genre Classification MATLAB
Classified of any audio song file into a musical genre (Rock, Pop, Classical, Jazz, and Metal) using only the low-level features
(~83% accuracy)
Used unsupervised learning methods K-Means, Self-Organization Maps (SOM)
PUBLICATIONS
Senthilnath, J., Karnwal, Nitin & Sai Teja, D. (2014). “Crop type classification based on clonal selection algorithm for high
resolution satellite image.” International Journal of Image, Graphics and Signal Processing, MECS, Vol. 6(9), pp. 11-19