Teradata: The Analytic Platform Provider
by Ron Powell
Originally published March 8, 2018
Patrick, there’s obviously a lot of excitement around the Teradata Analytics Platform. Can you tell us what is new with the platform?
Patrick Deglon: We are announcing a new platform. It is basically taking the existing Teradata database and augmenting it with new features. A few years ago we added temporal, a different type of time series. Here we’re adding a new engine and new functions. We really are moving the company from a data warehouse company to an analytic platform provider. It is a great new change in the company.
How does that fit into Teradata’s overall strategy?
Patrick Deglon: Teradata has always been about pushing the limits of analytics, even at the time when a gigabyte was an insane amount of data. I spent eight years at eBay pushing the limits of analytics with Teradata, doing crazy joins and crazy analysis. Here we’re just pushing it a bit further. We’re trying to embrace the new deep learning and artificial intelligence (AI), as well as new languages like R and Python. TensorFlow is a massive phenomenon around the world that enables any data scientist to do deep learning and create the same model that Google created for recognizing a kitty, a horse or a car. And now we can do that on Python and other languages using TensorFlow. So we have to embrace this – it’s reality. We’re creating this platform where in the future we will have TensorFlow-type of engines, and those will enable data scientists to run Python and R code together with an analytics platform.
Is this part of the vision that Oliver has with the Sentient Enterprise?
Patrick Deglon: It is the platform for the Sentient Enterprise. Every single step fits nicely. The most interesting and complex step – the steps are not in a specific order, but it’s the fifth one – is autonomous decision making. It is, I think, the most interesting one.
We are living at a very interesting moment in time. In the 1700s, there was the first industrial revolution where steam-powered machines totally transformed the face of mankind – trains, automation in manufacturing, etc. One-hundred years later came the invention of electricity that put lights in the streets and created even more automation. Henry Ford is a great entrepreneur from that time. One-hundred years later with chips and electronics, we created computers – a massive revolution. Actually, most of Teradata’s success is about the digitalization of the world and how you can make better decisions by collecting all the available data.
We are entering the fourth industrial revolution: the rise of cyber physical systems such as the self-driving car. Or you walk in the store and you pick what you want and you don’t check out. The store recognizes you, and you actually pay with your online wallet. This is the cyber physical system in the enterprise – because every single company will do it – but they won’t necessarily create the self-driving car. What they’re looking at is the self-driving employee. In the old times, there was computer-assisted design. Maybe we’re looking for the computer-assisted employee. This is about enabling employees, teams and companies to make better, faster, autonomous decisions. To do that, it really is a convergence of robots and machine learning with people, jobs and business rules. And the two together is exactly what we’re building. We’re trying to build the platform that will enable machine learning and more traditional business rules to enable enterprises to accelerate their business execution.
From a customer perspective, what is the value of the Teradata Analytics Platform?
Patrick Deglon: If you are an existing Teradata customer, it’s the extreme scalability, robustness and user manageability of Teradata that incorporates all the capabilities and functions of Aster.
Does the Teradata Analytics Platform support multiple deployment options like the cloud or on premises?
Patrick DeGlon: Excellent question. The most interesting point is this analytics platform is not a new product. There is no product ID that exists. If you have an existing Teradata database and your software license is up to date – let’s say 16.20 and above, you have the analytics platform. You have it. We are following exactly the same step with Teradata Everywhere whether it’s on premises, on your hardware/software, or in the cloud. You can pick and choose any deployment option you want. They can select the cloud first. But if the cloud is too slow or does not work for some reason, you can always migrate back to on premises. It’s a very beautiful model.
So no lock-in?
Patrick DeGlon: No lock in – exactly.
Patrick, it’s been a pleasure talking to you about the Teradata Analytics Platform.
Recent articles by Ron Powell
Copyright 2004 — 2018. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC