Resume
7+ years of engineering leadership and project management, and 10+ years of software development experience in AI, Machine Learning and Data Engineering. Experience in scaling engineering orgs, fund raising for company, milestones setting, liaisons with partners. Hands-on technical mentor. Highly collaborative team player. For comprehensive resume, please message me on LinkedIn.
Industry Experience :
2017-Present
Engineering Leader (Software)
Engineering leader leading software team to deliver high quality AI and SaaS products (Full stack) and research software. Experience with scaling organizations and hitting challenging milestones.
2010-2017
Individual Contributor/Tech Lead (Software)
Individual contributor/tech lead working on software services and data backend systems.
Education :
2008-2013
University of Southern California
Ph.D. in Electrical Engineering and Computer ScienceGPA: 4.00/4.00
Dissertation: Novel Variations of Sparse Representation Techniques with Applications
Academic Emphasis: Artificial Intelligence, Machine Learning, Natural Language Processing, Machine Learning, Big Data, Convex Optimization, Compressive Sensing, Statistics, Speech Recognition, Algorithms
2006-2008
Stanford University
M.S. in Electrical Engineering and Computer ScienceGPA: 3.93/4.00
Academic Emphasis: Information Theory, Communications, Algebraic Coding, Convex Optimization, Signal Processing
2004-2006
University of Michigan, Ann Arbor
B.S.E. in Electrical Engineering and Computer Science, MathematicsGPA: 3.93/4.00
Academic Emphasis: Signal Processing, Controls Theory, Numerical Methods, Number Theory
2003-2004
Stanford University
Education Program for Gifted Youths (EPGY)GPA: 4.15/4.00
Academic Emphasis: Pure Mathematics
Publications (as primary author/contributor) :
Journal articles :
Novel variations of group sparse regularization techniques with applications to noise robust automatic speech recognition
IEEE Transactions on Audio, Speech, and Language Processing, Volume: 20, Issue: 4, May 2012.
Enhanced sparse imputation techniques for a robust speech recognition front-end
IEEE Transactions on Audio, Speech, and Language Processing, Volume: 19, Issue: 8, Nov 2011.
Conference papers :
Combining window predictions efficiently - A new imputation approach for noise robust Automatic Speech Recognition
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2013.
Combining window predictions efficiently - A new imputation approach for noise robust Automatic Speech Recognition
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2013.
Automatic speech recognition system channel modeling
Eleventh Annual Conference of the International Speech Communication Association, Sept 2010.
Engineering Expertise :
Specialty Areas
Artificial Intelligence, Machine Learning, Natural Language Processing, Algorithms design, Systems design, Fullstack - Backend and Frontend, Mobile, Research and Development
Programming Languages
C#, C++, C, Python, Perl, PhP, Swift, Objective-C, Java, Javascript, Matlab, HTML, CSS
Tools
AWS, GCP, Databricks, SQL/NoSQL, HPC, CVX, SciPy, Scikit-learn, GDAL, Pandas, Keras, Tensorflow, PyTorch, Spark, Plotly/Dash, ZeroMQ, RabbitMQ, Django, Node, Electron
Special Skills
I am one of the few people in the world who can solve all official World Cube Association (WCA) Rubik's puzzles blindfolded consistently. Enjoy the following videos!
This is a video of me rescrambling a Square-1 blindfolded. I am the only documented person in the world to achieve this. Rescrambling a puzzle involves randomly mixing up a puzzle, and then copying the configuration on another copy of the puzzle. And blindfolded of course means executing without looking.
Watch on YoutubeThis is a video of me solving all official World Cube Association (WCA) puzzles blindfolded. The puzzles are (in increasing difficulty) the 2x2 Rubik's Cube, Pyraminx, Clock, 3x3 Rubik's Cube, Skewb, 4x4 Rubik's Cube, 5x5 Rubik's Cube, 6x6 Rubik's Cube, 7x7 Rubik's Cube, Megaminx, Square-1
Watch on YoutubeThis is video of me solving the 3x3 Rubik's Cube blindfolded in 34 seconds (time including memorization).
Watch on YoutubeThis is a video of me solving the 3x3 Rubik's Cube behind my back in 48 seconds (time including memorization).
Watch on YoutubeThis is a video of me solving the 7x7 Rubik's Cube blindfolded. To watch the uncut version of the solving phase, click here.
Watch on YoutubeThis is a video of me solving the Megaminx (12-sided Rubik's Cube) blindfolded. To watch the uncut version of the video, click here.
Watch on Youtube