I am an Assistant Professor in the Computer Science Department at Carnegie Mellon University, with a courtesy appointment in the ECE department. I lead the TheSys research group at CMU, and also a part of the Parallel Data Lab (PDL). My research interests lie in the broad area of computer and networked systems with a current focus on reliability, availability, scalability, and performance challenges in data storage and caching systems, in systems for machine learning and in live video streaming.

A bulk of my past research has focussed on the storage/caching layer and in part on the application (specifically, machine learning) layer:

• Storage/caching: My research focus here has been on fault tolerance, scalability, load balancing, and reducing latency in large-scale distributed data storage and caching systems. We designed coding theory based solutions that we showed are provably optimal. We also built systems and evaluated them on Facebook's data-analytics cluster and on Amazon EC2 showing significant benefits over the state-of-the-art. Our solutions are now a part of Apache Hadoop 3.0 and are also being considered by several companies such as NetApp and Cisco.


• Machine learning: My research focus here has been on the generalization performance of a class of learning algorithms that are widely used for ranking. We designed an algorithm building on top of Multiple Additive Regression Trees, and through empirical evaluation on real-world datasets showed significant improvement over classification, regression, and ranking tasks. The new algorithm that we proposed is now deployed in production in Microsoft's data-analysis toolbox which powers the Azure Machine Learning product.

TheSys (theory + systems) research group

My research group is called "TheSys (Theory + Systems)" since we take a principled and holistic approach towards solving real-world problems considering both theoretical and systems perspectives. We design solutions rooted in fundamental theory as well as build systems that employ the resulting insights and solutions to advance the state-of-the-art.

Here is our group's Github.

I am fortunate to be advising and working with the following amazing students at CMU.

PhD students:
Jack Kosaian
Michael Rudow
Saurabh Kadekodi (co-advised with Prof. Greg Ganger)
Juncheng (Jason) Yang
Francisco Maturana

Devdeep Ray (Prof. Srini Seshan's PhD student)

Masters students:
Jiaan Dai
Jiaqi Zuo
Jiongtao Ye
Sai Kiriti Badam
Xuren Zhou

Undergraduate students:
Eliot Robson (CMU)
Ian Chiu (CMU)

Publications

(On Google Scholar)

Preprints

---

• "Parity Models: A General Framework for Coding-Based Resilience in ML Inference"
  
  Jack Kosaian, K.V. Rashmi, and Shivaram Venkataraman
  Available on arXiv, May 2019.
  


• "Learning a Code: Machine Learning for Approximate Non-Linear Coded Computation"
  
  Jack Kosaian, K.V. Rashmi, and Shivaram Venkataraman
  Available on arXiv, June 2018.
  Code available on Github.

Conference Papers

---

• "Optimizing video upload for time-shifted viewing of social live streams"
  
  Devdeep Ray, Jack Kosaian, K. V. Rashmi, and Srini Seshan
  To appear in SIGCOMM, 2019.
  
  
• "Cluster storage systems gotta have HeART: improving storage efficiency by exploiting disk-reliability heterogeneity"
  
  Saurabh Kadekodi, K. V. Rashmi, and Greg Ganger
  USENIX FAST, 2019.
  
  
• "Streaming Codes for Variable-Size Arrivals"
  
  Michael Rudow and K. V. Rashmi
  Allerton Conference on Communication, Control, and Computing, 2018.
  
  
• "EC-Cache: Load-Balanced, Low-Latency Cluster Caching with Online Erasure Coding"
  
  K. V. Rashmi, Mosharaf Chowdhury, Jack Kosaian, Ion Stoica, and Kannan Ramchandran
  USENIX OSDI, 2016.
  (Slides) (Talk video)


• "Optimal Systematic Distributed Storage Codes with Fast Encoding"
  
  Preetum Nakkiran, K. V. Rashmi and Kannan Ramchandran
  IEEE International Symposium on Information Theory (ISIT), 2016.


• "Having Your Cake and Eating It Too: Jointly Optimal Erasure Codes for I/O, Storage, and Network-bandwidth"
  
  K. V. Rashmi, Preetum Nakkiran, Jingyan Wang, Nihar B. Shah, and Kannan Ramchandran
  USENIX FAST, Feb 2015.
  Picked as the best paper of USENIX FAST 2015 by StorageMojo.


• "DART: Dropouts meet Multiple Additive Regression Trees"
  
  K. V. Rashmi, and Ran Gilad-Bachrach
  International Conference on Artificial Intelligence and Statistics (AISTATS), Feb 2015.


• A "Hitchhiker's" Guide to Fast and Efficient Data Reconstruction in Erasure-coded Data Centers
  
  K. V. Rashmi, Nihar B. Shah, Dikang Gu, Hairong Kuang, Dhruba Borthakur, and Kannan Ramchandran
  ACM SIGCOMM, Aug 2014.
  (Slides)
  Press: The Morning Paper
  Press: StorageMojo


• "One Extra Bit of Download Ensures Perfectly Private Information Retrieval"
  
  Nihar B. Shah, K. V. Rashmi and Kannan Ramchandran
  IEEE International Symposium on Information Theory (ISIT), 2014.


• "Fundamental Limits on Communication for Oblivious Updates in Storage Networks"
  
  Preetum Nakkiran, Nihar B. Shah, and K. V. Rashmi
  IEEE GLOBECOM, 2014.


• "A Piggybacking Design Framework for Read-and Download-efficient Distributed Storage Codes"
  
  K. V. Rashmi*, Nihar B. Shah* and Kannan Ramchandran
  IEEE International Symposium on Information Theory (ISIT), July 2013.
  (Slides)


• "Secure network coding for distributed secret sharing with low communication cost"
  
  Nihar B. Shah, K. V. Rashmi, and Kannan Ramchandran
  IEEE International Symposium on Information Theory (ISIT), July 2013.
  (Slides)    (Poster)


• "Regenerating Codes for Errors and Erasures in Distributed Storage"
  
  K. V. Rashmi*, Nihar B. Shah*, Kannan Ramchandran, and P. Vijay Kumar
  IEEE International Symposium on Information Theory (ISIT), July 2012.
  (Slides)


• "Information-theoretically Secure Regenerating Codes for Distributed Storage"
  
  Nihar B. Shah*, K. V. Rashmi* and P. Vijay Kumar
  IEEE Globecom, Houston, Dec. 2011.


• "Enabling Node Repair in Any Erasure Code for Distributed Storage"
  
  K. V. Rashmi*, Nihar B. Shah* and P. Vijay Kumar
  IEEE International Symposium on Information Theory (ISIT), St. Petersburg, 2011.


• "A Flexible Class of Regenerating Codes for Distributed Storage"
  
  Nihar B. Shah*, K. V. Rashmi* and P. Vijay Kumar
  IEEE International Symposium on Information Theory (ISIT), Austin, June 2010.


• "Explicit and Optimal Exact-Regenerating Codes for the Minimum-Bandwidth Point in Distributed Storage"
  
  K. V. Rashmi*, Nihar B. Shah*, P. Vijay Kumar and Kannan Ramchandran
  IEEE International Symposium on Information Theory (ISIT), Austin, June 2010.


• "Explicit Codes Minimizing Repair Bandwidth for Distributed Storage"
  
  Nihar B. Shah*, K. V. Rashmi*, P. Vijay Kumar and Kannan Ramchandran
  IEEE Information Theory Workshop (ITW), Cairo, Jan.2010.


• "Explicit Construction of Optimal Exact Regenerating Codes for Distributed Storage"
  
  K. V. Rashmi*, Nihar B. Shah*, P. Vijay Kumar and Kannan Ramchandran
  Allerton Conf., Urbana-Champaign, Sep.2009.

Workshop Papers

---

• "A Solution to the Network Challenges of Data Recovery in Erasure-coded Distributed Storage Systems: A Study on the Facebook Warehouse Cluster"
  
  K. V. Rashmi, Nihar B. Shah, Dikang Gu, Hairong Kuang, Dhruba Borthakur, and Kannan Ramchandran
  USENIX HotStorage, June 2013.
  (Slides)    (Video)

Bio

Rashmi K. Vinayak is an assistant professor in the Computer Science department at Carnegie Mellon University. She recieved her PhD in the EECS department at UC Berkeley in 2016, and was a postdoctoral researcher at AMPLab/RISELab and BLISS. Her dissertation received the Eli Jury Award 2016 from the EECS department at UC Berkeley for outstanding achievement in the area of systems, communications, control, or signal processing. Rashmi is the recipient of the IEEE Data Storage Best Paper and Best Student Paper Awards for the years 2011/2012. She is also a recipient of the Facebook Fellowship 2012-13, the Microsoft Research PhD Fellowship 2013-15, and the Google Anita Borg Memorial Scholarship 2015-16. Her research interests lie broadly in the area of computer and networked systems with a current focus on addressing reliability, availability, scalability, and performance challenges in data-storage-and-analytics systems and live video streaming based on theoretical foundations.