Neeraja J. Yadwadkar

DSC_0190 2

I am a post-doctoral researcher in the Computer Science Department at Stanford University. My research interests are in Cloud Computing, Computer Systems, and Machine Learning. I graduated with a PhD in Computer Science from University of California, Berkeley. My dissertation was on Automatic Resource Management in the Datacenter and the Cloud. I received my masters in Computer Science from the Indian Institute of Science, Bangalore, India.

Most of my research straddles the boundaries of systems, and Machine Learning (ML). Advances in Systems, Machine Learning (ML), and hardware architectures are about to launch a new era in which we can use the entire cloud as a computer. New ML techniques are being developed for solving complex resource management problems in systems. Similarly, systems research is getting influenced by properties of emerging ML algorithms, and evolving hardware architectures. Bridging these complementary fields, my research focuses on using and developing ML techniques for systems, and building systems for ML.

neeraja emails

Publications

  • [NEW] Llama: A Heterogeneous & Serverless Framework for Auto-Tuning Video Analytics Pipelines
    Francisco Romero*, Mark Zhao*, Neeraja J. Yadwadkar, and Christos Kozyrakis
    (in submission)

  • [NEW] SmartHarvest: Harvesting Idle CPUs Safely and Efficiently in the Cloud
    Yawen Wang, Kapil Arya, Marios Kogias, Manohar Vanga, Aditya Bhandari, Neeraja J. Yadwadkar, Siddhartha Sen, Sameh Elnikety, Christos Kozyrakis, and Ricardo Bianchini
    (to appear in EuroSys 2021)

  • [NEW] Practical Scheduling for Real-World Serverless Computing
    Kostis Kaffes, Neeraja J. Yadwadkar, and Christos Kozyrakis
    (revision for EuroSys 2021)

  • [NEW] What Serverless Computing Is and Should Become: The Next Phase of Cloud Computing
    Johann Schleier-Smith, Vikram Sreekanti, Anurag Khandelwal, Joao Carreira, Neeraja J. Yadwadkar, Raluca Ada Popa, Joseph E. Gonzalez, Ion Stoica, and David A. Patterson
    (contributed article to appear in Communications of the ACM)

  • INFaaS: A Model-less Inference Serving System
    Francisco Romero*, Qian Li*, Neeraja J. Yadwadkar, and Christos Kozyrakis
    (in submission)

  • Centralized Core-granular Scheduling for Serverless Functions
    Kostis Kaffes, Neeraja J. Yadwadkar, and Christos Kozyrakis
    The ACM Symposium on Cloud Computing 2019 (SoCC), November 2019

  • A Case for Managed and Model-less Inference Serving
    Neeraja J. Yadwadkar, Francisco Romero, Qian Li, and Christos Kozyrakis
    The 17th Workshop on Hot Topics in Operating Systems (HotOS), May 2019

  • Cloud Programming Simplified: A Berkeley View on Serverless Computing
    Eric Jonas, Johann Schleier-Smith, Vikram Sreekanti, Chia-Che Tsai, Anurag Khandelwal, Qifan Pu, Vaishaal Shankar, Joao Carreira, Karl Krauth, Neeraja Yadwadkar, Joseph E Gonzalez, Raluca Ada Popa, Ion Stoica, David A Patterson
    EECS Department, University of California, Berkeley Technical Report, February 2019

  • Context: The Missing Piece in the Machine Learning Lifecycle
    Rolando Garcia, Vikram Sreekanti, Neeraja J. Yadwadkar, Daniel Crankshaw, Joseph E. Gonzalez, Joseph M. Hellerstein
    Workshop on Common Model Infrastructure at KDD, 2018

  • Selecting the Best VM across Multiple Public Clouds: A Data-Driven Performance Modeling Approach
    Neeraja J. Yadwadkar, Bharath Hariharan, Joseph E. Gonzalez, Burton Smith, and Randy Katz
    ACM Symposium on Cloud Computing (SoCC), 2017

  • Multi-Task Learning for Straggler Avoiding Predictive Job Scheduling
    Neeraja J. Yadwadkar, Bharath Hariharan, Joseph E. Gonzalez, and Randy Katz
    Journal of Machine Learning Research (JMLR), 2016

  • Faster Jobs in Distributed Data Processing using Multi-Task Learning
    Neeraja J. Yadwadkar, Bharath Hariharan, Joseph E. Gonzalez, and Randy Katz
    SIAM International Conference on Data Mining (SDM), 2015

  • Wrangler: Predictable and Faster Jobs using Fewer Resources
    Neeraja J. Yadwadkar
    , Ganesh Ananthanarayanan, and Randy Katz
    ACM Symposium on Cloud Computing (SoCC), 2014

  • Discovery of Application Workloads from Network File Traces
    Neeraja J. Yadwadkar
    , Chiranjib Bhattacharyya, K. Gopinath, Thirumale Niranjan, and Sai Susarla
    Usenix Conference on File and Storage Technologies (FAST), 2010

Teaching

  • EE282: Computer Systems Architecture, Winter 2019
    Co-Instructor with Prof. John Hennessy 

  • CS162: Operating Systems and Systems Programming, Fall 2017
    Graduate Student Instructor with Prof. Ion Stoica 
    Also, taught a couple of lectures (and it was a lot of fun!):
    - Caching (Finished), Demand Paging [slides: pptx pdf]
    - General I/O [slides: pptx pdf]

  • CS162: Operating Systems and Systems Programming, Spring 2013
    Graduate Student Instructor with Prof. Anthony Joseph

Academic Service

  • Diversity and Inclusion Chair, SoCC 2021

  • Co-Organizer, MLArchSys: Workshop on ML for Computer Architecture and Systems, ISCA 2021

  • Program Committee, WORDS: Workshop On Resource Disaggregation and Serverless, ASPLOS 2021

  • Area Chair - SysML (Systems for ML and ML for Systems), Journal of Systems Research (JSys), 2021

  • Area Co-Chair - Serverless Computing, Journal of Systems Research (JSys), 2021

  • Co-Founder, Journal of Systems Research (JSys), 2020

  • Program Committee, Conference on Machine Learning and Systems (MLSys), 2021

  • External Reviewer, ASPLOS 2021

  • Program Committee, SoCC 2020

  • Program Committee, Workshop on Machine Learning for Systems at NeurIPS 2020

  • Reviewer, ICML 2020

  • Program Committee, Systems and Machine Learning Conference (SysML), 2020

  • Poster Co-chair, Poster Program Committee, SoCC 2019

  • Program Committee, SoCC 2019

  • Program Committee, HotCloud 2019

  • Program Committee, Workshop on Resource Disaggregation (WORD), ASPLOS 2019

  • Reviewer, ICML 2019

  • Reviewer, NeurIPS 2019

  • Reviewer, NeurIPS 2018

  • External reviewer, OSDI 2016

  • External reviewer, Usenix ATC 2015

  • External reviewer, FAST 2014

Talks

  • Machine Learning for Resource Management
    Chalmers AI Research Centre, Chalmers University of Technology, Gothenberg, Sweden, May 2019 

  • A Case for Managed and Model-less Inference Serving
    The 17th workshop on Hot Topics in Operating Systems (HotOS'19), Bertinoro, Italy, May 15th, 2019 

  • Model-based Resource Allocation in the Public Cloud
    Platforms Lab Seminar, Stanford, CA, January 2019 

  • Machine Learning for resource management in Distributed Systems
    Invited Speaker at Workshop on ML for Systems at NeurIPS 2018, December 8th, 2018 

  • Machine Learning for Resource Management in the Datacenter and the Cloud
    Lawrence Berkeley National Lab, Berkeley, CA, January 2018 

  • Machine Learning for Resource Management in the Datacenter and the Cloud
    Platforms Lab, Stanford, CA, November 2017 

  • Machine Learning for Resource Management in the Datacenter and the Cloud
    Microsoft Research, Redmond, WA, November 2017 

  • Selecting the Best VM across Multiple Public Clouds: A Data-Driven Performance Modeling Approach
    ACM Symposium on Cloud Computing (SoCC), Santa Clara, CA, September 2017 

  • Selecting the Best VM across Multiple Public Clouds using PARIS: A Data-Driven Performance Modeling Approach
    RISELab/VMware Day, Berkeley, CA, May 2017 

  • Selecting the Best VM across Multiple Public Clouds using PARIS: A Data-Driven Performance Modeling Approach
    Google, Mountain View, CA, May 2017 

  • Data-Driven Modeling for Cloud-Hosted Systems' Management and Optimization
    Smule, San Francisco, CA, Jan 2017 

  • Let your Workloads Choose your VMs in the Cloud using PARIS
    RISELab Winter Retreat, Berkeley, CA, Jan 2017 

  • Data-Driven Modeling for Cloud Management and Optimization
    Splunk, San Francisco, CA, July 2016 

  • Data-Driven Modeling for System Management and Optimization
    SAP Dublin, CA, June 2016 

  • PARIS: Model Based Performance Estimation Across the Cloud
    AMPLab Summer Retreat June 2016 

  • Managing Sample Bias in a Model-Based Cluster Resource Manager
    AMPLab Summer Retreat June 2016 

  • The Judgement of PARIS: Performance-Aware Resource Inference System
    Microsoft Research, Redmond, Intern Talk, August 2015 and AMPLab Winter Retreat, January 2016 

  • Faster Jobs in Distributed Processing Systems using Machine Learning
    Department Seminar, Department of Computer Science and Automation (CSA), Indian Institute of Science (IISc), May 2015 

  • Faster Jobs in Distributed Data Processing using Multi-Task Learning
    SIAM International Conference on Data Mining (SDM), April 2015 

  • Wrangler: Predictable and Faster Jobs using Fewer Resources
    ACM Symposium on Cloud Computing (SoCC), November 2014 

  • Wrangler: A Machine Learning Approach for Straggler Avoidance
    AMPLab Summer Retreat, May 2014 and AMPLab All Hands 2014  

  • Zone Localization Methods and Services
    Software Defined Buildings (SDB) Winter Retreat, Jan 2014 

  • Discovery of Application Workloads from Network File Traces 
    Usenix Conference on File and Storage Technologies (FAST) Feb 2010 and Riverbed Technology, Feb 2010 

News