Teradata Reaches for the Sky with Cloud Turnaround
- By Paul Mah
- October 12, 2023
More than 65% of data science projects in large organizations fail, according to Steve McMillan, the CEO of Teradata. He was in town in August for the Singapore leg of the firm’s Possible 2023 conference.
“They don't fail because they didn't have a great idea, or they weren't able to develop the model. They tend to fail because it's very difficult to put that model in production based on real enterprise data and real technology,” he said.
Optimized for AI
Everyone is talking about AI today, courtesy of OpenAI and its groundbreaking breakthroughs in generative AI. But taking advantage of AI is easier said than done, notes McMillan, which is an argument I can agree with.
For instance, only a few participants at a recent roundtable discussion about AI that I moderated say they are piloting AI projects. According to McMillan, the challenge organizations face with successfully rolling out AI projects lies with their data.
“Is the data of high quality, can you trust the data, is the integrity of the data good? Only with trusted data can AI truly be successful and make the right recommendations.”
“I think a lot of organizations talk about you know, generative AI and AI in theory, but they don't have the capabilities to actually put it into production,” he said.
Teradata has invested heavily in integrating with platforms such as AWS SageMaker to help enterprises realize their AI initiatives: “We have developed a model ops capability in the Teradata platform that can help our customers take their [AI models] from an idea or concept around a question that needs to be answered and take it all the way through to production.”
“Our focus is to help organizations harmonize all of the data in their environment. And utilize that data to feed data science models, large language models, AI capabilities to get real business outcomes.”
Losing the imagination of the industry
But why are competitors getting more of the thunder despite the long list of banks, airlines, and enterprises that rely on Teradata products? In fact, a delegate I spoke to over lunch ticked off multiple banks in Singapore that currently use Teradata.
When asked, McMillan was brutally honest about the missteps of the 40-year-old firm: “We hadn't invested in our cloud strategy. We didn't embrace the cloud as the way forward.”
Teradata “lost the imagination” of the industry, he said, as he highlighted Teradata’s cloud and on-premises revenues to illustrate the dichotomy. “We had maybe 50 million dollars of recurring revenue in the cloud. And one and a half billion dollars of recurring revenue on-premise.”
Switching things around meant pouring a lot more resources into the cloud: “It was 20% investment in cloud and 80% in on-premises solutions previously. We inverted that and took all those fantastic Teradata technologies and made it available in the cloud.”
“We have really only been on this journey for the last three years. But in those three years, we has made some tremendous progress. We now have hundreds of our customers in the cloud.”
Putting customers first
The team has since completely rearchitected Teradata as a cloud-native solution available on both AWS and Azure. Indeed, over 100 engineers who left Teradata have come back – they loved the company and rejoined Teradata when the company adopted a cloud-first strategy.
For now, McMillan says he wants to help his customers get the most out of the data no matter where it is, utilizing Teradata’s finely-tuned engine to get real business outcomes – no duplication of data required.
“No matter if it's in a native object store, in AWS S3, blob storage, or on-premise, we will open up our ecosystem to be able to access all of that data. And our customers can utilize Teradata to do real queries with real results for real business outcomes.”
Teradata is on track to achieve half a billion dollars of recurrent revenue in the cloud by end-2023, he says. And by 2025, the company is aiming to have over a billion dollars of recurring revenue in the cloud.
The Teradata today is completely different from what it was, asserts McMillan. “We have completely transformed the company and our capability and it starts with the technology.”
Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].
Image credit: iStockphoto/Boonyachoat
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.