Tech Lead and Manager, MapReduce Team Quantcast
THIS JOB HAS EXPIRED
We seek a technical lead and manager to lead the research and development of our map-reduce platform and other large-scale execution engines.
Here are a few examples of the type of work we do in the Cluster R&D group:
QFS, our petabyte-scale file system 2x more efficient and faster than HDFS. (https://github.com/quantcast/qfs/wiki/Performance-Comparison-to-HDFS)
Quantsort, a revolutionary map-reduce sort architecture that speeds up map-reduce significantly and simplifies greatly the management of shared map-reduce clusters. (https://www.quantcast.com/inside-quantcast/2012/10/quantsort-revolution-in-map-reduce-performance-and-operation/)
Turbo Sawzall, a fast analytics platform that scans billions of records in seconds. (https://www.quantcast.com/inside-quantcast/2011/11/turbo-mapreduce/)
A candidate for this position will have proved interest in, and experience with, big data, distributed systems, availability and reliability of large systems, fault-tolerance, efficiency and performance.
We offer you the opportunity to make an impact and grow in a fast paced environment at the bleeding edge of commodity cluster computing. We need your passion, drive, energy and experience with distributed or supercomputing systems.
Opportunities:
Rare R&D environment. Quantcast?s business model relies heavily on big data. This makes for a healthy R&D environment where ideas are driven by and can be validated quickly at scale on realistic workloads. Moreover, we develop the whole cluster software stack, including a high-performance distributed file system, which means you won?t be constrained by arbitrary limitations of third party software.
Make an impact. Quantcast produces world-class analytics and advertising platforms with a relatively small engineering team. Your impact will be much higher here than at a large company.
Short feedback cycle. Our development cycles are short. You get to see your contribution used and appreciated quickly.
Great timing. We are a fast growing pre-IPO technocracy. It?s a rush, don?t miss it.
Responsibilities:
Drive and manage the design and development of cluster computing. We?re running one of the largest mapreduce clusters on thousands of cores, petabytes of data, and terabits of networking. We?re looking at both improving traditional map-reduce for today?s workload and hardware, and at new processing paradigms for fast analytics.
Lead, don?t follow. Drive innovation and contribute to intellectual property.
Requirements:
Passionate, self-driven
Experience leading a high performance computing team or project
Extensive development experience in UNIX systems programming and design
Excellent development skills in C++ or Java
Expert knowledge of distributed or supercomputing systems
Excellent problem solving skills, including system level debugging
About Quantcast
Quantcast is an audience measurement and targeting company. The pioneer of direct audience measurement, we start with the industry?s most in-depth understanding of digital audiences to help marketers and publishers buy and sell the most effective targeted advertising and drive conversions through the full funnel. Our products let publishers match their audience to the exact consumers an advertiser wants to reach with impression level targeting. Ranked Fast Company?s #3 Most Innovative Company on the Web and the Overall Winner of AlwaysOn?s Global 250 Top Private Companies, Quantcast is used by the world?s leading advertisers, the top 10 media agencies and 100+ million web destinations. Launched in 2006, Quantcast is headquartered in San Francisco and backed by Founders Fund, Polaris Venture Partners, Revolution Ventures and Cisco Systems.
| Location: |
400 Second Street
Suite 350
San Francisco, CA 94107
United States
|
THIS JOB HAS EXPIRED