I am a researcher in Azure AI Frameworks on rotation from Microsoft Research. Currently I am particularly interested in the intersection of reinforcement learning, and graph / kernel automatic scheduling for novel accelerators (GPUs, NPUs) and am exclusively focused on this topic.

Previously I was a researcher in the Machine Learning area at Microsoft Research, Redmond for 8 years.

I conduct not only fundamental research but also love to build high-quality software for e.g.,

  • Archai a PyTorch-based Neural Architecture Search framework. Models produced by Archai are used by millions worldwide every day and handle billions of queries.
  • AirSIM a photo-realistic simulator for robotics which is widely used by the community.

I finished my PhD at the Robotics Institute, Carnegie Mellon University. I do fundamental as well as applied research in machine learning, control and computer vision with applications to autonomous agents in general and robotics in particular. My interests include decison-making under uncertainty, reinforcement learning, artificial intelligence and machine learning. My work has been honored with Best Paper of the Year Shortlist at the International Journal of Robotics Research.

From 2015-2023 I used to regularly area-chair/review for NeurIPS, ICLR, ICML, AutoML. On occasion for ICRA, IROS, IJRR, JFR, CVPR, ECCV, ICCV, TMLR.


  • Kernel and Graph Schedule Search
  • Neural Architecture Search
  • AutoML
  • Reinforcement Learning
  • Robotics
  • Planning
  • Vision


  • PhD in Robotics, 2015

    Carnegie Mellon University

  • MS in Robotics, 2012

    Carnegie Mellon University

  • Bachelor of Electrical Engineering, 2007

    Delhi College of Engineering



Principal Researcher

Azure AI Frameworks

Aug 2023 – Present Redmond, Washington

Principal Researcher

Microsoft Research

Aug 2019 – Aug 2023 Redmond, Washington

Senior Researcher

Microsoft Research

Jul 2015 – Aug 2019 Redmond, Washington

PhD Student

Robotics Institute, Carnegie Mellon University

Jul 2010 – Jul 2015 Pittsburgh, Pennsylvania



Aditya Modi

University of Michigan, Summer 2018

Alex LaGrassa

CMU, Summer 2020

Angela Lin

University of Texas, Summer 2019

Artem Rozantsov

EPFL, Summer 2016

Benjamin Hepp

ETH Zurich, Summer 2017

Brian Axelrod

Stanford University, Summer 2016

Dilip Arumugam

Stanford University, Summer 2019

Elizabeth Bondi

Harvard University, Fall 2017

Felix Berkenkamp

ETH Zurich, Summer 2017

Francisco Garcia

University of Massachusetts, Fall 2016

Ganesh Jawahar

UBC, Summer 2021

Hanzhang Hu

CMU, Summer 2018

Khanh Nguyen

UMD, Summer 2018

Mike Roberts

Stanford University, Summer 2016, 2017

Mojan Javaheripi

UCSD, Summer 2021

Ramya Ramakrishnan

MIT, Summer 2017, 2018

Sanjiban Choudhury

CMU, Summer 2016

Shushman Choudhury

Stanford University, Summer 2020

Simon Ramstedt

MILA, Summer 2017

Tianle Cai

Princeton, Summer 2022

Wen Sun

CMU, Summer 2016

Yuhong Li

UIUC, Summer 2022

All Publications

Quickly discover relevant content by filtering publications.

What Makes Convolutional Models Great on Long Sequence Modeling?

Convolutional models have been widely used in multiple domains. However, most existing models only use local convolution, making the …

AutoDistil: Few-shot Task-agnostic Neural Architecture Search for Distilling Large Language Models

Knowledge distillation (KD) methods compress large models into smaller students with manually-designed student architectures given …

Efficient forward architecture search

We propose a neural architecture search (NAS) algorithm, Petridish, to iteratively add shortcut connections to existing network layers. …

Anytime neural networks via joint optimization of auxiliary losses

This work considers the trade-off between accuracy and test-time computational cost of deep neural networks (DNNs) via mph{anytime} …

Overcoming blind spots in the real world: Leveraging complementary abilities for joint execution

Simulators are being increasingly used to train agents before deploying them in real-world environments. While training in simulation …

Vision-based Navigation with Language-based Assistance via Imitation Learning with Indirect Intervention

We present Vision-based Navigation with Languagebased Assistance (VNLA), a grounded vision-language task where an agent with visual …

Discovering blind spots in reinforcement learning

Agents trained in simulation may make errors in the real world due to mismatches between training and execution environments. These …

Learn-to-score: Efficient 3d scene exploration by predicting view utility

Camera equipped drones are nowadays being used to explore large scenes and reconstruct detailed 3D maps. When free space in the scene …

Submodular trajectory optimization for aerial 3d scanning

Drones equipped with cameras are emerging as a powerful tool for large-scale aerial 3D scanning, but existing automatic flight planners …

Adaptive information gathering via imitation learning

In the adaptive information gathering problem, a policy is required to select an informative sensing location using the history of …

Learning to gather information via imitation

The budgeted information gathering problem - where a robot with a fixed fuel budget is required to maximize the amount of information …

Safety-aware algorithms for adversarial contextual bandit

In this work we study the safe sequential decision making problem under the setting of adversarial contextual bandits with sequential …

Airsim: High-fidelity visual and physical simulation for autonomous vehicles

Developing and testing algorithms for autonomous vehicles in real world is an expensive and time consuming process. Also, in order to …

Vision and learning for deliberative monocular cluttered flight

Cameras provide a rich source of information while being passive, cheap and lightweight for small and medium Unmanned Aerial Vehicles …

Predicting Sets and Lists: Theory and Practice

Increasingly, real world problems require multiple predictions while traditional supervised learning techniques focus on making a …

Predicting multiple structured visual interpretations

We present a simple approach for producing a small number of structured visual outputs which have high recall, for a variety of tasks …

Gauss Meets Canadian Traveler: Shortest-Path Problems with Correlated Natural Dynamics

In a variety of real world problems from robot navigation to logistics, agents face the challenge of path optimization on a graph with …

Knapsack constrained contextual submodular list prediction with application to multi-document summarization

Many prediction domains, such as ad placement, recommendation, trajectory prediction, and document summarization, require predicting a …

Learning monocular reactive uav control in cluttered natural environments

Autonomous navigation for large Unmanned Aerial Vehicles (UAVs) is fairly straight-forward, as expensive sensors and monitoring devices …

Classification of plant structures from uncalibrated image sequences

This paper demonstrates the feasibility of recovering fine-scale plant structure in 3D point clouds by leveraging recent advances in …

Contextual Sequence Prediction with Application to Control Library Optimization

Sequence optimization, where the items in a list are ordered to maximize some reward has many applications such as web advertisement …

Efficient Optimization of Control Libraries

A popular approach to high dimensional control problems in robotics uses a library of candidate “maneuvers” or “trajectories”. The …