amazee.ai launches private AI gateway for sensitive data

amazee.ai has launched a private AI gateway for region-specific deployments, aimed at organisations that handle sensitive data.

The managed gateway offers an OpenAI-compatible API that lets customers connect applications to multiple AI models, including GPT, Claude, Gemini and Kimi, while keeping requests within selected regions and applying zero data retention by default.

The service is aimed at sectors such as financial services, healthcare, government and legal services, where data handling rules and security requirements can limit the use of public AI services. amazee.ai says the gateway is designed so customer data is not retained and cannot be used by outside parties to train AI models.

Users access the service through a single endpoint that routes requests to different model providers. Because the gateway follows the OpenAI API format, customers can keep existing software development kits, frameworks and applications in place, according to amazee.ai.

Regional controls

A central element of the launch is the ability for customers to choose where workloads are processed. The gateway supports region pinning in the United States, the European Union's GDPR region, Australia, Germany, Switzerland and the UK, the company says.

That structure is likely to appeal to organisations that need to show where data is handled and stored, particularly when AI systems are used across borders. Companies in regulated sectors have faced growing scrutiny over whether prompts, documents and other inputs might leave approved jurisdictions or be retained by third-party model providers.

amazee.ai says it can also provide dedicated gateways for customers with more specific infrastructure needs, including deployments within a customer's own environment. That gives clients another option if internal policy or regulation prevents them from using a shared managed service.

Compliance focus

The service has ISO 27001 and SOC 2 Type II certification, according to the company. amazee.ai describes the gateway architecture as audit-ready for regulated environments, with private and isolated tenants for customers using the service.

The launch comes as businesses look for ways to use large language models without exposing internal information. While many AI providers now offer business terms on data use and retention, buyers in heavily regulated industries continue to seek tighter contractual and technical controls, along with greater certainty over geographic processing.

amazee.ai is part of Mirantis. The business sits alongside amazee.io, which Mirantis acquired in 2022.

Michael Schmid, Co-founder and Managing Director of amazee.ai, outlined the company's rationale for the product. "We're addressing the biggest issue in enterprise AI adoption - data governance of sensitive corporate data, along with unauthorized use, to maintain operational control and avoid data leakage," Schmid said.

He added: "We built the amazee.ai Private AI Gateway as an enterprise trust layer: an OpenAI-compatible gateway that gives organizations sovereignty, audit readiness, and stronger security controls by default."

Mirantis has focused on cloud infrastructure and open source software, with a customer base that includes large enterprises and service providers. By linking that background with AI infrastructure and model access, the group is seeking to position itself around the operational concerns that often slow adoption in larger organisations.

For buyers, the practical question is whether services such as this can reduce the friction of adopting generative AI tools without requiring major changes to existing development workflows. amazee.ai argues that compatibility with the OpenAI standard, combined with regional controls and no data retention, will let customers add model access while keeping tighter oversight of how information is handled.

The gateway includes multi-model routing, sovereign deployment controls and isolated tenancy, according to the company.