日本語 
DataRobot and Nebius announced a strategic partnership that combines DataRobot’s agent workforce platform, co-developed with NVIDIA, with an AI cloud infrastructure designed to bring enterprise-grade agents to production. This solution integrates Nebius’s dedicated NVIDIA GPU-based infrastructure with DataRobot’s platform to deliver a validated “AI factory.” This allows enterprises to avoid the operational complexities of traditional hyperscalers and seamlessly build, deploy, monitor, and govern AI agents. This solution can reduce the time to production deployment of AI agents from months to as little as days.
This partnership addresses a critical gap in real-world operations. Running AI agents requires sustained inference performance, runtime governance, and infrastructure that operates predictably as demand increases. Given the volatility of performance and costs, enterprises are increasingly valuing AI-specific infrastructure rather than generic, existing cloud solutions. By combining DataRobot’s agent workforce platform, Nebius’s dedicated AI cloud, and NVIDIA’s validated AI optimization stack with GPUs and open software, DataRobot and Nebius deliver scalable, high-performance agent-based AI that transcends the limitations of traditional clouds.
This collaboration will enable DataRobot to directly deploy NVIDIA GPU-driven inference workloads, such as custom models and agent execution services, onto Nebius-managed Kubernetes. As a result, agent intelligence and governance will be truly integrated with NVIDIA’s high-performance infrastructure built specifically for AI.
Also Read: Serverworks partners with Databricks on data analytics
By eliminating the performance degradation caused by the load of other users in a shared environment (noisy neighbor problem), a challenge in traditional cloud environments, this solution delivers overwhelming performance that maximizes the inherent power of physical servers, with the low latency and predictable throughput essential for always-on agents. Powered by NVIDIA Hopper and NVIDIA Blackwell generation data center GPUs and protected by NVIDIA NeMo Guardrails , this platform provides a clear path to running AI as a trusted, production-grade system within the enterprise. This co-designed stack delivers enterprise-grade functionality without the vendor lock-in, cost volatility, architectural limitations, and inflated costs often associated with traditional providers. This allows organizations to scale their agent workforce with absolute confidence.
SOURCE: PRTimes
English