日本語 
Cloudera K.K., a company providing AI for data everywhere, and Mistral, a global leader in AI research and development, have announced a strategic partnership to provide secure sovereign AI directly to enterprise data, wherever that data is located.
The partnership brings together Cloudera’s hybrid data and AI platform and Mistral’s sovereign AI models. Together, the companies will provide an environment where enterprises can build, customize, and operate AI using their own private data.
AI can be deployed directly where the data is stored. This includes cloud environments, on-premises systems, edge environments, sovereign environments, and even air-gapped environments. Sensitive information does not have to be moved outside the company’s control.
For companies that manage large amounts of sensitive and valuable data, this gives them more control over where and how AI is used. Enterprises can run inference in private environments, customize AI with their own data, and deploy AI applications within their existing security and governance frameworks.
They can also have more flexibility over AI costs as usage increases.
Abhas Ricky, Chief Business Officer and GM of Applied AI at Cloudera, said that enterprise AI is entering a new stage. According to him, companies need more than access to high-performance AI models. They also need the freedom to use their own data under their own conditions and create intelligence that is specific to their business.
The collaboration with Mistral will allow companies to run AI where their data is located, customize AI using their own intellectual property, and maintain control over their data, infrastructure, and costs. This combination of intelligence and control is intended to help companies move AI from experimentation into production.
Delivering Intelligence Where the Data Resides
For companies in highly regulated industries, moving sensitive or proprietary information to external AI services can create regulatory, security, and operational issues.
Cloudera and Mistral aim to address these issues by bringing secure sovereign AI directly into enterprise data environments.
Mistral’s AI models and tools will be integrated with Cloudera’s hybrid data and AI platform. Customers will be able to use up to 30 exabytes of data to run customized AI models and applications.
The deployment can be adapted to different environments. These include public cloud, private cloud, on-premises environments, and fully air-gapped systems. Enterprises can maintain consistent governance and control over their data and the context used by AI.
The partnership covers a broad range of Mistral’s AI models and tools. These include inference, chat, coding, document intelligence, and speech. This gives Cloudera customers more options for using AI with their enterprise data.
The partnership also gives enterprises more flexibility in managing AI costs. Companies can run inference within their own environments and choose the deployment model and infrastructure that best fits their workloads as usage grows. They do not have to rely entirely on pay-as-you-go public API models.
From Private Inference to Proprietary Intelligence
The partnership will also integrate Mistral Forge, which allows enterprises to build advanced AI models based on their own knowledge.
Companies using Cloudera’s platform will be able to use large amounts of private enterprise data within a managed environment to customize and train models.
For organizations with petabytes of proprietary data, this creates an opportunity to turn years or even decades of accumulated company data and industry knowledge into differentiated AI. At the same time, companies can retain ownership and control over both their data and the intelligence created from it.
Developers and other users can build AI-powered experiences using private enterprise data within their own environments. Sensitive business information and intellectual property do not need to be exposed externally.
These applications can include conversational access to governed data, AI-assisted software development, and agent-based workflows.
Cloudera and Mistral will also work together on next-generation edge AI. The companies aim to enable more advanced inference capabilities closer to where enterprise data is generated.
This could allow companies to use intelligent applications in environments where centralized cloud infrastructure is not practical. Examples include locations with limited network connectivity, low-latency requirements, or limited computing resources.
Kamal Brar, SVP of Partnerships and Alliances at Mistral, said the company’s mission is to help businesses use advanced AI without giving up control of their data, infrastructure, or intellectual property.
He added that Cloudera manages some of the world’s most valuable enterprise data assets, and the partnership provides an opportunity to bring Mistral’s technology directly to those data environments.
Through the collaboration, customers will be able to build AI that reflects their own data and expertise and deploy it securely wherever their business requires.
Yuji Yamaga, President and CEO of Cloudera, said that as Japanese companies continue to adopt AI, there is growing interest in using advanced AI while maintaining security, governance, and data sovereignty for critical information.
He said the partnership combines Cloudera’s hybrid data and AI platform with Mistral’s AI models and gives Japanese customers another option for using AI according to their own environments and requirements while keeping their data under their control.
Cloudera will continue to support customers in using their own data and expertise to create AI-driven innovation and new value in a secure and flexible way.
Expanding Enterprise AI Options
The partnership also expands the Cloudera Enterprise AI Ecosystem.
Through this ecosystem, Cloudera works with AI model providers, infrastructure companies, application providers, and technology companies to give customers more flexibility when building and deploying enterprise AI.
With Mistral joining the ecosystem, customers will be able to access and customize Mistral’s models and AI capabilities directly alongside their governed enterprise data.
Also Read: IIJ and NTT Docomo Business Demonstrate Data Sharing
The collaboration also strengthens the companies’ focus on an open approach to enterprise AI. Companies can choose from different models, infrastructure, and deployment environments instead of being tied to a single technology stack for their data and AI strategies.
The joint Cloudera and Mistral solution will be available through Cloudera’s enterprise sales team and partner ecosystem.
Additional integrations and features will be introduced in the future.
English