Datacast: Data & Analytics at Scale

You write a query and get data back. In between, magic happens. In this episode, we talk with Ron Luebke [https://www.linkedin.com/in/ron-luebke-b9078a177/] about the “query optimizer”—a critical component of that magic. We talk about query optimizers in general, what makes Teradata’s query optimizer a bit more magic than the rest, and how that magic gets extended beyond the relational database to other parts of a complete analytics ecosystem.

In this episode, Carrie Ballinger [https://www.linkedin.com/in/carrie-ballinger-79851050/] introduces us to a critical requirement of managing data and analytics at scale—workload management. * What is “workload management”? * Why and when workload management is necessary * The most important thing workload management enables * How is workload management managed? * Different types of workload management * How workload management improves overall throughput * Interesting workload management stories

“There are many types of data that have many different types of relationships, and you need to manage that data based on value.” — Rob Armstrong, Teradata This week, Rob Armstrong shares all you need to know about data management by value—in just two shapes. LINKS * The Circle and Square, All You Need to Know About Data and Analytics [https://www.teradata.com/Blogs/The-Circle-and-Square,-all-you-need-to-know-about] [blog post]

Our guest this week is Tim Miller, from Teradata’s Technology & Innovation Office. He was a principal in the development of the first commercial in-database data mining system, which makes him a perfect choice to provide an overviewof the data science process. New to data science? There may be occasional bouts of jargon you don’t understand. If so, don’t worry! Tim (naturally) uses terminology from the data science domain, but we always bring it back to plain-English explanations. * What is data science? * CRISP-DM:Cross-industry standard process for data mining [Wikipedia [https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining]] * Fundamental: What is the business problem we’re trying to address? * Data science is a team sport * Supervised and unsupervised learning * How do you decide which techniques and algorithms to use? * What is analytics ops? * Happiness is “in-platform analytics”

“There are things in every single company that should be done in the cloud. On the flip side of the coin, not everything belongs in the cloud.” — Marc Clark Not sure what data and analytics scenarios make the most sense to do in the cloud? Maybe more importantly, are you sure which don’t? Marc Clark, Teradata’s Director Global Sales Enablement and Strategy, shares some clarity on the cloud. What are the right use cases for cloud analytics? What are the wrong ones? What’s the best way for a company to dip their toes into cloudy waters?