Having undergone a modern makeover and unleashed its advanced analytics platform Vantage in the past six months, database and analytics provider Teradata has successfully transformed itself from an outdated data warehouse appliance company into a leading business data analytics company.
With almost four decades of experience in data, having watched the industry mature and the big data revolution sweep across industries, Teradata has a strong idea of where the industry is heading.
Timeline for Automation
- March 26, 2019
At the annual Analytics Universe conference, recently held in Las Vegas, Nevada, Verdict asked four of Teradata’s leading experts one question: What is the future of data analytics?
Best of breed technologies
Big data – the collection of large data sets used by technologies such as AI to make informed decisions – is a frequently used buzzword that most will have come across.
But according to Stephen Brobst, Teradata’s Chief Technology Officer, the tech industry isn’t interested in big data.
“I’ll tell you that at Silicon Valley we do not say big data strategy. It’s just data strategy. Big data is just part of the data,” Brobst told Verdict.
While the focus is currently on big data, Brobst believes that businesses need to combine both big data and small data “to get the full value”. The challenge that businesses face is finding the right tool to manage and analyse these different kinds of data.
Currently, many advanced analytics applications run on Apache Hadoop, an open source framework that handles data processing and storage. However, Brobst doesn’t believe that one piece of software should be viewed as the solution to all of the industry’s issues.
“Hadoop doesn’t solve all the problems, and now the hype has worn off and the reality has set in,” Brobst said. “Some think big data equals Hadoop. I don’t buy that. It’s a broader context.”
“There are lots of different technologies you can use to explore large unstructured data, and Hadoop is part of the answer, but it’s not the full credit answer.”
In the future, Brobst believes that businesses will begin to diversify, using niche, best-of-breed solutions and technologies that are best equipped to deal with specific kinds of data.
“The future is ecosystems, not silver bullets,” Brobst said. “You need to integrate best of breed technologies that each do their part of the job, and not think that one technology is gonna be the silver bullet.”
3 Things That Will Change the World Today
However, for those focused on the business side of data analytics, the future is whatever the company’s high-spending clients want. And according to Oliver Ratzesberger, Chief Operating Officer of Teradata, what they want is simplification.
Given the sheer scale of data and possibilities that it provides to improve business operation and performance, many will often make use of software from various vendors to manage their data and analytics operations.
“Nowadays I often talk to C-level executives that tell me: ‘I’m spending $2bn a year on data analytics. Couple thousand head count, couple hundred systems – it’s unsustainable. I can no longer as a company grow that, be agile with that. I need to find a way to simplify that. I need much, much fewer platforms. If I could, I want a single platform instead of 800’,” Ratzesberger told Verdict.
Offering “tens of billions of dollars a year of opportunity”, this is at least the future of Teradata, which has already started to deliver a solution with its Teradata Vantage data intelligence platform.
Released in October last year, Vantage offers this unity by allowing users to integrate a range of tools and languages, and offering flexibly across different data types, formations and stores. According to Teradata, it is the only software of its kind that is capable of “managing all of the data, all of the time”.
Companies like Teradata already help to automate the processing and storage of data. However, according to Atif Kureishy, Teradata’s Vice President of Global Emerging Practices, the future of big data and analytics will bring further automation.
“I think a lot of the trends are driving more automation into data science,” Kureishy told Verdict. “How do you really extract out the data scientists, to make it not so tedious and amorous on the data scientists, but instead using all the principles of AI [artificial intelligence] and math and statistics to essentially accelerate how you take in data, build learning models and deploy them?”
While a great deal of automation is involved in data handling and analytics, it still requires a great level of human development and monitoring to ensure that these automated systems run effectively.
“The capability to deliver end-to-end analytics comes down to having people with various skill sets that can really take something from data ingestion, doing the data wrangling, building the muild and then also automating the productionisation of that model,” explained Yasmeen Ahmad, Director at Think Big Analytics.
“So you need data engineers, you need DevOps specialists, software engineers, data scientists, business intelligence and cognitive design people to help you build a user experience around that, you might need apps consultants who can take the analytic models and feed them out into the enterprise.”
People in these various roles will still be required to build these analytic models. However, Ahmad believes that many of the time-consuming management and monitoring tasks will be removed.
“A lot of the work, it’s manual hand-cranking – let me have a look at this data set, if there’s data missing, is it poor quality?,” Ahmad told Verdict. “We’ve got AI, we’ve got algorithms that can now do that. They can automate that process.”
“it’s like applying AI to AI itself,” Kureishy said.
Blockchain: A future use?
While the current focus may be on AI and automation, Teradata keeps a close eye on other emerging technologies through its Global Emerging Practices division.
With blockchain often singled out as the next big thing to disrupt business, Verdict questioned Kureishy on the potential future use of the technology within the data analytics industry.
While he admitted that there was “nothing much beyond the concept”, Kureishy does see a potential in blockchain as a security solution.
“The sort of balance of machine intelligence that AI offers and then trust and confidentially, which is the other aspect of that – information insurance and information security – those things go hand in hand and not everyone understands that,”Kureishy said.
“In a future state, if you think about what’s happening in the enterprises, more and more data with more and more analytics means you’re going to have more and more machine intelligence.
“The only way that’s going to happen is if you have trust and assurance that what’s happening is what you intend, so that’s where all of the information security principles and confidentiality integrity come into play. And the only way to do that at scale is with something that’s distributed, such as blockchain.
“Though over time, as we get more and more proliferation of these types of intelligence happening in an enterprise, they’re going to figure out those things have to come together.”