Backslash Partners, a software development company, has unveiled Backslash AIOps 1.0, a platform designed to manage GPU resources and streamline AI development and operations. The product has been registered for TTA GS (Good Software) certification testing.
Built on Kubernetes, Backslash AIOps 1.0 enables organizations to manage GPU resource requests, approvals, partitioning and reclamation while supporting AI model training and deployment through MLOps capabilities.
Users can request GPU resources by specifying their intended purpose and required duration. Administrators can review and approve requests, after which resources are automatically reclaimed when the approved period ends. The platform also records request histories, approval records and GPU utilization data.
The platform supports GPU partitioning technologies including time-slicing, Multi-Process Service (MPS) and Multi-Instance GPU (MIG). This allows organizations to allocate GPU resources according to workload requirements and operational environments.
Backslash AIOps 1.0 also provides GPU resource allocation, scheduling and monitoring capabilities. It is designed to help organizations improve resource utilization, reduce idle GPU capacity and manage allocation policies more efficiently.
For AI development workflows, the platform supports integration with JupyterLab for development environments and Argo Workflows for workflow orchestration. Monitoring data and resource usage records can be collected to support operational visibility and management.
The platform also incorporates access control, multi-tenant resource management and audit logging to help organizations manage GPU infrastructure in shared environments.
Backslash Partners plans to continue developing its GPU resource management solutions and expand its business in the AI infrastructure and MLOps markets.
The company stated that Backslash AIOps 1.0 was registered as a product for TTA GS certification testing, marking a step toward strengthening the product’s reliability and market readiness.