Company

O’Reilly Data Show:
Peter Bailis on Machine Learning and Operational Analytics

By Grant Shirk - October 11, 2019

This week, our CEO and founder Peter Bailis joined Ben Lorica on the O’Reilly Data Show podcast for a conversation about how machine learning can be applied to the problem of operational analytics in the enterprise. 

The two covered a wide range of topics, from Peter’s research at Stanford University to why we believe that today’s enterprises can no longer rely on outdated analytics tools to inform their daily decisions. The full podcast is embedded below, or you can see the notes and transcript here on the Data Show page

 

 

Here’s a quick summary of the key topics:

  • The quest to build a better transaction database, and the resulting “wild ride” for the NoSQL movement
  • Why it looks like we’re in a Golden Age of Data, but most businesses are missing out
  • Why most of the research into deep learning isn’t applicable to the enterprise 
  • How machine learning and better query models can improve operational analytics and business intelligence
  • The fundamental inequality between the rate at which metrics change and the rate data changes
  • Machine learning benchmarks, including DAWNBench and MLPerf
  • And finally, trends in data management and tools for machine learning development, governance, and operations

 

Ben is the Chief Data Scientist at O’Reilly and the Program Director for O’Reilly’s Strata Data and AI Conferences.  

Read more about how Sisu helps businesses use all of their structured data in real time to inform daily decisions across the enterprise.

 


Read more

September 3, 2019

Sisu at the O'Reilly AI Conference San Jose

Peter presented on Usable Machine Learning at the 2019 O’Reilly AI conference in San Jose.

Read more

June 27, 2019

Three Design Principles for Operational Analytics

With more data, we need faster tools for analysis. At Sisu, we've identified three key design principles for making analytics accessible and understandable.

Read more