About Me

Current :

I am a Normal Senior Principal Machine Learning Scientist working on Statistics, Machine Learning, Artificial Intelligence, Big Data, Reinforcement Learning, GenAI - add other buzz words here -  making history at the one and only Amazon. I lead a small team in the Learned Systems Group, within the Database, Analytics and ML Org in AWS, working on products such as Amazon Redshift and Amazon RDS. Here is a recent paper from SIGMOD 2023 on some of the ML for Systems work we've done over the last few years.

Some recent services I pretend to, used to, will or sometimes even do help with - with high variance and low mean levels of usefulness (yes, that means my usefulness is often negative) - include Amazon Personalize, Amazon Forecast, Amazon SageMaker RL, Amazon Redshift and Amazon RDS. I also work on fun stuff for services like AWS Lambda and with Rashmi Gangadhariah on Artificially Intelligent chatbots and ML for UX design. I am pretty proud that some of my work for Amazon Personalize was recognized as the best paper in the Applied Data Science Track at KDD 2020. I led the science teams that launched Amazon Redshift ML, DevOpsGuru for RDS and GuardDuty RDS Protection. Most recently I led science teams that launched AI driven scaling and optimization in Redshift Serverless and Q for Generative SQL a text to sql agent for Amazon Redshift.

These and many other projects in our lab are hiring scientists and engineers. If you are interested, do reach out. 

My research interests lie at the intersection of AI, optimization, learning and inference particularly using them to understand, model and combat noise and uncertainty in real world applications. I'm particularly interested in Reinforcement Learning in practice and at scale. Broadly, I work on using ideas from online algorithms, optimization under uncertainty, control theory, game theory, artificial intelligence, graphical models and estimation theory to solve important problems at Amazon scale. My goal is to develop novel, somewhat theoretically well motivated optimization algorithms that `work'. My role model is Hari Seldon, though results in my time would be nice. For technical details on stuff I managed to sneak past peer review, see my publications page. 

Trajectory :

From 2014-2015 I was a post-doc at UC San Diego, in the Machine Learning Group with Prof. Charles Elkan and MESL, headed by Prof. Rajesh Gupta.

From 2011-2013 I was a Research Staff Member at IBM Research, India Research Lab in Bangalore where I worked on Next Generation Systems. I was a part of the High Performance Computing (HPC) group and also worked with the Smarter Energy group.

At the same time, I had the pleasure of working with the amazing game theory lab run by Prof. Narahari at the Indian Institute of Science. My collaborators included Shweta Jain, Swaprava Nath and Pankaj Dayama.

Before IBM (note the obfuscation of a start date), I was a Doctoral student in the Electrical and Computer Engineering department at Carnegie Mellon University, from where I graduated in 2011. I am fortunate to be advised (then and now) by Profs. Raj Reddy, Rohit Negi and Pradeep Khosla. I was also lucky to be mentored by Yaron Rachlin. My PhD research interests were in establishing theoretical and algorithmic limits and guarantees for large scale detection applications (think sensor networks) using ideas from information and coding theory.

In the past I have worked in robust speech recognition, text-independent speaker identification, speech science, machine learning and auditory scene analysis with Prof. Richard Stern. I also did some work on target tracking with Ryan Kerekes and Iris recognition (the body part not the flower) with Ryan Kerekes and Jason Thronton (who even in this day and age does not have a webpage).

At CMU I was associated with CENSIR, the Robust Speech Group and the Sphinx Speech Group.

I am a proud recipient of the National Talent Search (NTSE) and the Jawaharlal Nehru (JNCASR) scholarships from the the government of India during my undergraduate studies.

Service :

TPC of BuildSys 2016

Reviewer for   NIPS 2016

Very limited reviewer for ICML 2016 

Organizing MPE 2013+ Workshop on Data-aware Energy Use

TPC of E-Energy 2015 and DEN 2015

TPC of ECML PKDD 2014 and 2013

TPC of HotPower 2014 

Publicity Chair for  BuildSys 2014

Reviewer for the Journal of Climate

Reviewer for the Transactions on Cloud Computing

Education :  (kinda)

B.E - Electronics and Communication Engineering - M.S.R.I.T, Bangalore under VTU (03)

M.S - Electrical and Computer Engineering - Carnegie Mellon University, Pittsburgh(05)

PhD - Electrical and Computer Engineering - Carnegie Mellon University, Pittsburgh(11)

Disclaimer : The postings on this site are my own and don't necessarily represent Amazon's positions, strategies or opinions.