VESSL AI has shifted its primary business direction toward its GPU cloud platform, VESSL Cloud, aiming to simplify access to large-scale GPU resources for AI development, particularly for what it terms 'Physical AI.' This repositioning was highlighted during its exhibition at NVIDIA GTC 2026, where the company demonstrated its platform's capabilities for efficiently deploying AI development environments and managing substantial GPU needs without requiring clients to build their own hardware. The company's offerings include features like 'VESSL Run' for automated model training, 'VESSL Serve' for real-time deployment, 'VESSL Pipelines' for streamlining workflows, and 'VESSL Cluster' for optimizing GPU utilization within cluster settings.
The company has recently secured $12 million in funding, with reports of potential GPU cost reductions of up to 80% for its users. This financial backing supports its operational expansion and the further development of its machine learning operations (MLOps) platform. VESSL AI reports having over 2,000 users and a customer base of 50 enterprises, including notable names such as Hyundai, LIG Nex1, and TMAP Mobility.
Read More: New Smart Pillow Lets You Control Audio Without Touching Your Phone
VESSL AI’s platform is designed to address the complexities of developing and utilizing machine learning tools. The company, founded in 2020 by individuals with prior experience at entities like Google and AI startups, offers a suite of tools aimed at automating AI model training, streamlining data preprocessing, and optimizing GPU resource allocation. Its approach includes leveraging a multi-cloud strategy and the use of spot instances to navigate GPU availability and development costs.
The company has also established strategic partnerships, including collaborations with Oracle and Google Cloud in the United States. This infrastructure support appears to be crucial for their operations, as indicated by their engagement with Samsung SDS for GPU-as-a-Service (GPUaaS) through the Samsung Cloud Platform (SCP). This partnership allowed VESSL AI to offload infrastructure management, thereby enabling a greater focus on developing its MLOps and LLMOps platform services and exploring new business avenues. The SCP GPUaaS is noted for providing stable, on-demand GPU infrastructure with technical support, aiming to reduce engineering hours spent on troubleshooting and management.
Read More: LLM Performance Plateau Means Less Big Jumps, More Small Helps
The technical integration of VESSL AI’s services is facilitated through a Python package installable via pip (pip install vessl), with further guidance available through quickstart guides and example models on VESSL Hub. The company’s move towards a dedicated GPU cloud platform and its stated cost-saving capabilities place it within a competitive landscape that includes established cloud providers and other MLOps startups.