About the Author

andrei_ppDr. Andrei Khurshudov is a Director of Advanced IoT Analytics at Caterpillar. He is managing several teams of data scientists and software developers working on advanced monitoring and analytics solutions for the company’s fleet of connected machines and IoT devices.

Prior to this role, Andrei served as a Chief Data Officer at Formulus Black, a computer software startup from Jersey City, NJ.  Formulus Black develops sophisticated software for In-Memory Computing (IMC).

In the most recent role before Formulus Black, Andrei was a Chief Data Officer and a CTO at Alchemy IoT, an IoT startup from Boulder, Colorado.

Before AlchemyIoT, Andrei was a Chief Technologist at Seagate and focused on Big Data Analytics and Insights.

And prior to Seagate, Andrei worked at IBM, Hitachi, and Samsung (all in San Jose, California).

Andrei is an experienced executive leader with a long track record of building and managing best-in-class organizations and achieving measurable results. Andrei has extensive expertise in Big Data analytics and machine learning, IoT technology and analytics, mathematical modeling, financial and business analysis and forecasting, data storage technology (from storage components to Cloud data centers), distributed analytics (Hadoop, Spark), and product quality/reliability.

You can follow Andrei’s @siliconfuture on Twitter.

"Scorpion Corps" by Anya BerlovaAnya Berlova is the editor of blackboxparadox.com and almost every article published here is reviewed and approved by her.

Anya is a talented writer and an artist.

This is Anya’s website:  anyaberlova.com

Andrei’s recent and selected past publications and presentations:

  • CIO Review paper on how to build the best possible big data analytics organization:  here
  • CIO Review paper on Black Box Paradox:  here
  • My presentation a HCI workshop at the University of Minnesota, April 30th, 2019: Hyper-Converged Infrastructure:  Big Data and IoT opportunities and challenges. 
  • Invited talk at the 26th ASME annual conference on information and storage and processing systems (ISPS 2017) held at Hilton San Francisco District, San Francisco, California, USA from August 29–30, 2017: Short Introduction to Big Data Analytics, the Internet of Things, and their synergies.
  • CIO Review paper on using big data analytics to produce high-quality big data storage: here
  • These are links to Seagate’s Hadoop on Lustre/Hadoop Workflow Accelerator work (developed by my team):
    • Hadoop Workflow Accelerator white paper: here
    • Storage Review article: here
    • The Register article: here
    • HPC Wire article: here
  • Our presentation on Cloud Gazer – the data center health monitoring solution – presented at the Open Compute Summit in 2015: link. Cloud Gazer is designed to monitor and control health and performance of storage devices at the data center (drives, servers, server components) and detect and predict their failures.
  • My presentation at the Predictive Analytics conference in Chicago, November 2015: link
  • My paper with Dan Lingenfelter and Dimitar Vlassarev on modeling disk drive performance:  link
  • Invited talk at Symposium on Magnetic Storage Tribology and Reliability, Miami, Florida, 2008: Reliability of Solid State Drives.
  • Storage Visions Conference, Las Vegas, 2007:  Long-Term Data Storage.
  • ISPS Conference 2007: Future Information Growth and Storage Device Reliability.
  • My white paper from Seagate on how the workload impacts HDD reliability and data center total cost of ownership (TCO).  Just before this publication, I came up with the concept of a new workload-based HDD reliability specification or “workload rate limit” (WRL, measured in TB/year) and, after some internal work, Seagate has accepted it as a new specification now used to define different classes of products.  Eventually, the industry (WDC, HGST, Toshiba-Fujitsu) followed this example and accepted this entire concept as well: link
  • Seagate even came up with a clever video to illustrate the above concept:

  • Paper on Hard Disk Drive Reliability Modeling and Failure Prediction from 2007.  I think we were the first to propose and publish the physical model linking HDD reliability to relative humidity inside the drive.  The model was based on the idea of water condensation in the high-pressure areas under the slider, which resulted in a loss of supporting pressure and the reduction in effective slide-disk clearance:  link
  • Another disk drive reliability paper I wrote with George Tyndalllink
  • Paper on the monolayer disk lubrication and its impact on slider-disk clearance.  Do you know that the lubricant used to protect magnetic disks is just about 10 A thick? And 10 A = 1 nm, which equals 1E-9 m…  link
  • And this is the link to my Google Scholar page:  here

I don’t have my earlier papers as PDFs…  need to digitize them!

  • And this is my book from 2001:  they still have them on Amazon.


Contact Andrei