Getty Images/iStockphoto

Teradata integrates with Google Vertex AI

The vendors' new integration provides choices to customers who prefer Google as their cloud provider. It also helps Google present itself as more focused on enterprises.

Teradata and Google Cloud on Tuesday unveiled an integration of Google Cloud's Vertex AI machine learning platform with Teradata VantageCloud and ClearScape Analytics. The integration is aimed at helping organizations move from the AI experimentation phase to implementing AI into their workloads.

VantageCloud is Teradata's cloud-native data management and analytics platform. ClearScape Analytics is Teradata's business intelligence platform that lets users work with data in a database.

Integrations with multiple cloud providers

With the new integration, the vendors said that with VantageCloud, ClearScape Analytics and Vertex AI, Teradata users can build and train AI and machine learning models faster; integrate datasets from separate environments, data lakes and object stores; and operationalize Vertex AI models at scale in VantageCloud.

This partnership is not the first between the vendors. In 2019, Google partnered with Teradata to help Google offer Teradata data analytics products on Google Cloud Platform. In addition, Google users could also run Teradata's cloud analytics platform (then known as Vantage).

The new integration of Vertex AI with VantageCloud and ClearScape Analytics is similar to Teradata's recent integrations with other cloud providers. In March, Teradata integrated with Azure Machine Learning. The longtime data management and analytics vendor also has a similar arrangement with the Amazon SageMaker ML platform.

The moves to integrate with different cloud providers let Teradata meet customers on the cloud platforms they're using, said Mike Leone, an analyst at TechTarget's Enterprise Strategy Group.

They're probably interested in Vertex because they can see that it addresses particular challenges, such as [being reusable], being able to scale easily, being able to track experiments, being able to manage and administer this process of being able to monitor adaptive models.
Donald FarmerFounder and principal, TreeHive Strategy

"It's really customer flexibility and openness," he said. "It's not so much, 'Oh, what I can do in Vertex [that] I can't do in AWS SageMaker?' It's just a matter of customers asking for access to AI tools on their terms." That may be with AWS or with Google Vertex AI.

Moreover, Teradata's specific market focuses on bigger, sophisticated enterprises that demand trustworthiness and reliability, said Donald Farmer, founder and principal at TreeHive Strategy.

"They're probably interested in Vertex because they can see that it addresses particular challenges, such as [being reusable], being able to scale easily, being able to track experiments, being able to manage and administer this process, and being able to monitor adaptive models," Farmer said.

Teradata's customers will also be drawn to Vertex AI because of its pre-trained AI models, autoML capabilities, reliability and managed services that deliver scalability, Leone said.

For Google's part, the integration helps it make Vertex AI a better and more capable product, he added.

Moreover, with Google seemingly falling behind Microsoft in the AI race, this is a chance for Google to stand out and show it doing something different, Farmer said.

"They're not just building cool stuff; they're building enterprise-ready AI technology," he said. "A good way of showing that they are enterprise ready is through a partnership with an enterprise company like Teradata."

Some challenges

However, the Vertex AI integration will only make working with AI and ML tools easier for some of Teradata's customers.

Some customers are still early in the AI process and will need guidance to embrace and get the most value from Vertex AI, Leone said. However, this is a good opportunity for Teradata to distinguish itself from other data management vendors.

"They're going to have to fight that battle a little to prove it out," Leone added. "Even though they have customers that are using AI, even though they know full well that they can deliver AI at scale, this is going to give them a more robust story."

Another challenge is data governance and ensuring that private and personally identifiable customer data is secure, Farmer said.

"How can you train a model to give interesting and useful insights to the enterprise but still be able to guarantee that that data remains private, secure and well governed?" he said.

Esther Ajao is a news writer covering artificial intelligence software and systems.

Dig Deeper on Machine learning platforms

Business Analytics
Data Management