Teradata CEO: 'ChatGPT is a great catalyst for analytics innovation'
'We're in the sweet spot for data, analytics and cloud', says Steve McMillan
'We're in the sweet spot,' said Steve McMillan, CEO of Teradata on a recent visit to London for the company's 'Possible' event.
"If you look at the top three spending priorities for CIOs, it's security number one, then cloud computing, with data and analytics at number three. We're in the sweet spot of data and analytics and cloud."
It wasn't always thus. In fact, 40-year-old Teradata only ventured wholeheartedly into the cloud three years ago with its Vantage platform, which runs on Azure, AWS and Google clouds as well as in data centres.
However, public cloud income now makes up almost a third of the company's annual recurring revenues at $414 million, up 77% year-on-year.
McMillan insists that while Teradata's focus is on cloud, all new innovations will continue to apply to data centre versions too, avoiding the dichotomies that can make on-premises users of some enterprise software feel like poor relations. He cited the company's partnership with Dell, announced in May, as its evidence of its commitment to the hybrid model.
"We've done a lot of engineering work so that the Teradata platform can run on Dell's converged infrastructure at the same kind of performance levels as our core appliance. So it's really about a choice of deployment model."
Sixty percent of Teradata customers that have gone to the cloud have retained an on-premises capability, McMillan said. "CIOs want to be in control."
Teradata's hybrid and multi-cloud game plan
The hybrid and multi-cloud models are very much part of the company's game plan, which in a nutshell is to be extensible, open and to run anywhere.
The hybrid multi-cloud trend is being driven by several factors, a key one of which is regulation. The finance sector has always been restricted as to where it can store sensitive data, but the Bank of England and the Prudential Regulation Authority recently announced new regulations for important banking and financial services industries (FSIs) around "stressed exits".
FSIs must be able to demonstrate the ability to "substitute the service provider or bring the outsourced service back in-house, including estimated costs, operational impact, risks, and timeframe of an exit in stressed and non-stressed scenarios," according to the regulation.
In other words, UK banks must have viable hybrid and multi-cloud capabilities by law. Other nations will likely follow.
Elsewhere, organisations are simply seeing the wisdom of separating duties, McMillan said, speaking of a customer that has "front office in AWS, back office in Azure, Teradata in both.
"They want flexibility from a negotiation perspective, and they recognise that some of the services that are available in Azure are better than in AWS, and vice versa. I think that's what a really modern hybrid cloud environment starts to look like."
Because of the size and complexity of workloads, and the time, effort and expense involved in moving them, data has "gravity". Teradata's pitch here, and the focus of recent patents, is to acknowledge that gravity and take the analytics to the data, not the other way around.
"Organisations just aren't going to have all the data in one place, from now on," McMillan explained. Teradata analytics instances "respect the data gravity," by being able to run "pushdown queries" between platforms, he added.
"We're not moving the data around. We're allowing the customer to use that data no matter where it is. We push the query to the data, rather than pulling the data to the query. I think that's fairly unique."
AI - a great catalyst
As well as burnishing its hybrid multi-cloud credentials, Teradata, like every other company in this field, is keen to promote its AI competencies. Once again, the company has found itself in a serendipitous sweet spot, according to McMillan.
"We announced our ClearScape analytics capabilities in August 2022, then ChatGPT was announced in November 2022. We were pretty excited about that. It was great in terms of the global interest and it's a great catalyst for new innovation from a data and analytics perspective."
Teradata is well positioned to address concerns over the trustworthiness of LLMs, by managing data lineage and tracking decisions back to the input, he insisted.
The company integrates with the large language models offered by the cloud platforms, allowing customers to develop their own models based on production data. Several are already using this capability, including The Very Group, which is optimising promotions and supply chains in this way.
"I think we'll see more and more of that," McMillan said. "Supply chain optimisation, improving developer productivity, and so on. We're also using it to improve our products. We're building our product and integration capabilities to continuously enhance how we do math at scale and large language models. So, super exciting, in terms of the opportunities that it's going to unlock as we move forward."