Data modeling and annotation solution provider Sama launched new integrations with cloud hyperscalers Microsoft and Google aimed at accelerating customer onboarding of data and lowering the costs of doing so.

Customers looking to migrate 2,000 Tb of data from Google Cloud or Microsoft Azure can expect to save nearly $400,000 by transferring files from the customer’s cloud Server to a Server of the same architecture and in the same region – and spend a fraction of the traditional time needed to migrate large amounts of data.

Immense datasets are required for running artificial intelligence (AI) models at edge or public cloud computing sites. “Sama’s new capability with major cloud providers allows us to access these datasets on behalf of our clients quickly, securely and cost-effectively,” Sama SVP of Product and Technology Duncan Curtis said.

Sama’s new multicloud integrations also reduce the onboarding time to just one day, according to the vendor. That accelerated timeline limits the need for Sama to design custom coding solutions to access customer servers and data.

The vendor’s custom data pipeline integration with technologies like Python Software Development Kits (SDKs) and the Databricks platform, for example, allows Sama to pair with existing data flows and reduce the need for customer engineers to build annotation and validation workflows.

“The faster we can complete onboarding, the faster we can assist our clients in getting their models to market,” Curtis said.

Sama claims to have baked security into these multicloud integrations. Preventing breaches is a top priority – especially considering that large datasets potentially include personally identifiable information (PII).

To that point, Sama does not retain copies of its customers’ private data; instead, it accesses customer files where they exist and applies any modifications before writing the updated file directly back to the customer’s Server. This strategy helps meet IT compliance requirements and enhances data security.

The Google Cloud and Microsoft Azure integrations are a piece of Sama’s larger efforts to speed up time-to-market and lower the overall costs of AI model ownership. Sama boasts a 99% client acceptance rate for AI training data and an annotation delivery rate of more than 300 million frames, more than 850 million shapes and 10 billion annotation points each month.

This multicloud integration also points to a growing embrace of multicloud operating models.

The move comes as hyperscalers like Google Cloud, Microsoft Azure and Amazon Web Services (AWS) are falling under scrutiny in regard to anticompetitive practices that discourage the use of competing public cloud services.

Google Cloud notably waived a portion of its data transfer fees earlier this month in an effort to silence anticompetitive accusations, and this integration is the latest foray toward an open cloud ecosystem.