Juan A. Colmenares
e-mail: juancol [at] eecs [dot] berkeley [dot] edu
Welcome to my personal page.
I am a staff software engineer at LinkedIn.
I currently work on a new distributed graph database for LinkedIn's Economic Graph,
designed to serve declarative queries at high throughput and with low latency.
From 2012 to 2018, I was a researcher at
Samsung Research America (SRA) and
the tech lead of the Smart Systems Team since 2016.
In 2013, I was an industrial researcher for Samsung in the
Ubiquitous Swarm Lab at UC Berkeley.
Previously, I was a post-doctoral scholar in
the UC Berkeley's Parallel Computing Laboratory (Par Lab), working with
Prof. John D. Kubiatowicz,
Prof. Krste Asanović, and
Prof. David Wessel.
I completed my Ph.D. in the EECS Department at UC Irvine,
where I had the privilege of working under the supervision of Prof. Kwang-Hae (Kane) Kim.
Before doing my Ph.D., I was a faculty member in the Engineering School's Applied Computing Institute at the
University of Zulia, in Venezuela.
I served as its director in 2003-2004.
My CV is here
(outdated, of course).
Post-doctoral Scholar. University of California, Berkeley. (September 2009 - July 2012)
Mentor: Prof. John D. Kubiatowicz.
Ph.D. in Electrical and Computer Engineering. University of California, Irvine. (March 2009)
Dissertation: Derivation of Service-time Bounds of Methods in Time-triggered Message-triggered Objects.
Advisor: Prof. K. H. (Kane) Kim
M.Sc. in Applied Computing. University of Zulia, Venezuela. (November 2001)
Thesis: Development of an Efficient Bayesian Global Optimization Algorithm and its Integration into a Distributed Processing Environment.
Advisor: Prof. Salvador Pintos
Electrical Engineer. University of Zulia, Venezuela. (December 1997)
Ranking: #1 among 58 graduates.
- Scalable and latency-sensitive systems
- Data management and analytics
- Real-time distributed computing systems
- Adaptive resource centric computing
Selected Publications (Full List)
A single-node datastore for high-velocity multidimensional sensor data
IEEE Big Data 2017
[doi] | Extended Preprint [arXiv]
Power consumption models for multi-tenant server infrastructures
ACM TACO 14(4) 2017 [doi]
KV-Cache: A scalable high-performance web-object caching for manycore (Runner-up for the Best Paper Award)
Our technical report "A Scalable
High-Performance In-Memory Key-Value Cache using a Microkernel-Based Design"
contains more details and an extensive evaluation of KV-Cache.
Juggle: Addressing extrinsic load imbalances in SPMD applications on multicore computers
Cluster Computing 2013
Tessellation: Refactoring the OS around explicit resource containers with continuous adaptation
Our technical report
"Building an Adaptive Operating System for Predictability and Efficiency"
presents additional details and a more recent evaluation of our prototype.
Ingestion, indexing and retrieval of high-velocity multidimensional sensor data on a single node.
Presented in the Seminar Series of the CS Department at the University of California, Irvine.
Irvine, CA, USA. January 12, 2018.
Will computer systems with performance guarantees ever go mainstream? (keynote)
Presented at the 15th IEEE International Symposium on High Assurance Systems Engineering (HASE 2014).
Miami, FL, USA. January 10, 2014.
Refactoring the OS around explicit resource containers with continuous adaptation.
Presented at the Par Lab's End of Project Celebration.
Hertz Hall. University of California at Berkeley.
Berkeley, CA, USA. May 30, 2013.
Summer Student Mentoring
- 2016: Alexander Adams (Cornell) and Ying Lu (USC)
- 2015: Younghwan Go (KAIST)
- 2014: Edgar Pek (UIUC)
- 2013: Gage Eads (UC Berkeley)
Last update: October 2018.