Neoclouds, MSPs, sovereign AI providers, and IT service companies are under mounting pressure to deploy AI-ready environments quickly, onboard tenants on demand, and maintain strict uptime and isolation—all without months of custom integration. Spectro Cloud and Netris are tackling that challenge head-on with a new integrated solution combining PaletteAI and Netris NAAM (Network Automation, Abstraction, and Multi-Tenancy).

The joint platform lets teams provision tenant-ready AI environments with coordinated changes across cluster, workload, and network layers in a single, automated workflow. The result: faster time-to-revenue, instant safe onboarding, lower operational risk, and a future-ready architecture for NVIDIA GPU ecosystems.

From Months to Days: Key Benefits

• Faster Time-to-Revenue: Pre-validated full-stack deployments—from bare metal to billable tenant—replace months of manual assembly.

• Instant, Safe Tenant Onboarding: Hard isolation enforced at the network layer with automated tenant provisioning, no cross-team tickets required.

• Lower Operational Risk: API-driven orchestration ties network changes directly to cluster lifecycle, reducing outages and strengthening SLAs.

• Multi-Fabric Consistency: Unified control plane across Ethernet, DPUs, InfiniBand, and NVLink fabrics.

• Secure Multi-Tenancy: Netris DPU agents enable NVIDIA DOCA HBN for network isolation and concurrent multi-tenant workloads on bare-metal GPU hosts.

• Future-Ready Architecture: Validated today on NVIDIA BlueField-3 DPUs and designed for upcoming NVIDIA BlueField-4 and Vera Rubin architectures.

How the Integration Works

• PaletteAI: Delivers curated, deployable blueprints across OS, Kubernetes, storage, and the NVIDIA AI Enterprise ecosystem with governance and self-service for AI teams.

• Netris NAAM: Automates network orchestration, enforces hard isolation, and removes manual ticket-driven bottlenecks.

The integration ensures end-to-end tenant orchestration: PaletteAI deploys validated platform blueprints, while Netris provisions matching network constructs. Day-two operations—expansions, onboarding, decommissioning—trigger automatic network updates via Netris APIs. Compliance-aligned, air-gapped deployments are supported for regulated environments.

Built on the NVIDIA Ecosystem

The joint solution is validated across NVIDIA Spectrum-X Ethernet, Quantum-X InfiniBand, BlueField DPUs, NVLink, and the NVIDIA AI Enterprise software stack. By unifying infrastructure, networking, and AI workloads, the platform enables operators to move from infrastructure to inference in days, not months.

Executive Perspectives

“Building AI infrastructure at scale has always required stitching together cluster and network manually. Spectro Cloud and Netris eliminate that—meeting at the DPU to deliver a production-ready AI factory,” said Alex Saroyan, CEO and Co-Founder of Netris.

“The next generation of cloud infrastructure is the AI factory. PaletteAI automates full-stack management, and Netris extends that to network orchestration and hard multi-tenancy. Together, we deliver a turnkey platform for operators to move from bare metal to revenue quickly,” said Tenry Fu, CEO of Spectro Cloud.

This solution positions operators to standardize deployments, accelerate production readiness, and confidently scale multi-tenant AI workloads, transforming GPU infrastructure into a fully automated “AI factory.”

Get in touch with our MarTech Experts.