What Rafay Systems does

Rafay Systems, Inc. is a privately held infrastructure orchestration and workflow automation company founded in 2017 and headquartered in Sunnyvale, California. The company sells the Rafay Platform, which sits atop open-source Kubernetes, GPU Operator, Prometheus, Argo, Terraform, and related cloud-native components, and adds proprietary middleware (Rafay controller, workflow engine, metering engine, PaaS Studio SKU Designer, and Token Factory) to deliver self-service, multi-tenant GPU and Kubernetes infrastructure across public cloud, private data center, sovereign, air-gapped, and edge environments. The platform serves large enterprises (Freddie Mac, Samsung, Amgen, Verizon, Blackstone), neoclouds, sovereign AI cloud operators (TELUS, Indosat, Ooredoo, Cassava Technologies, Yotta, Firmus, AI-GDC), telecommunications providers, and government entities (U.S. Air Force), with notable operational outcomes including a 92% reduction in platform team headcount at Freddie Mac and 28-customer onboarding in under six weeks at Indosat. Revenue is generated through node or cluster-based subscription licensing with optional usage-based consumption, professional services for implementation, and OEM/white-label licensing arrangements, with annual recurring revenue of approximately $25.1 million in 2024 across roughly 158 employees.

Go-to-market is enterprise-led via direct field sales, supplemented by systems integrator partners (Accenture, Deloitte) and hardware OEM channel (Dell, Cisco, Penguin Solutions). The company raised approximately $33 million in disclosed funding (an $8M Series A in 2019 and a $25M Series B led by ForgePoint Capital in August 2021), and has been recognized by Gartner as a 2023 Cool Vendor in Container Management and by GigaOm as a 2025 Leader and Outperformer in Kubernetes Management. In 2026 Rafay achieved NVIDIA AI Cloud-Ready validation and launched Token Factory for token-metered AI service monetization, with named AI factory deployments across six continents.

Name

Rafay Systems

Legal name

Rafay Systems, Inc.

Website

https://rafay.co

Company type

Private

Founded year

2017

Operating status

Operating

Headcount range

11–50 employees

Short description

Rafay Systems provides infrastructure orchestration and workflow automation software that transforms GPU and Kubernetes infrastructure into self-service, multi-tenant AI and cloud-native platforms for enterprises, neoclouds, sovereign AI operators, and telecommunications providers across six continents.

Ownership category

akta.pro rank

Product category

Cloud Infrastructure Orchestration

NAICS

Software Publishers (513210), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Computer Systems Design and Related Services (54151)

SIC

Services-Prepackaged Software (7372), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)

akta.pro secondary industries

AI Compute Cloud & GPU-as-a-Service (HDAAAAAK), AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ), Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC), AI Observability, Monitoring & Evaluation Platforms (Drift, Quality, Safety) (HDAEANAF), End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)

Keywords

• Kubernetes orchestration platform
• GPU infrastructure management
• AI cloud orchestration
• Multi-tenant PaaS software
• Infrastructure automation platform

Headquarters

HQ city

Sunnyvale

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Marketing or Sales, Operations, Infrastructure

Revenue model

1. Platform Subscription Licensing: Node-based or cluster-based subscription model. Pricing varies by deployment model (SaaS, self-hosted, air-gapped), scale, support requirements, and use cases. Volume discounts available as node count increases. Enterprise Support available at additional 20% of subscription fee. EDU/GOV discounts available.
2. GPU-as-a-Service / AI Service Monetization: Token-metered AI service delivery where organizations expose AI models via APIs and charge based on token consumption rather than raw GPU hours. Enables model-as-a-service, inference-as-a-service, and AI marketplace business models.
3. Professional Services: Implementation services including platform deployment, environment onboarding, SKU configuration, AI service operationalization, and Build-Validate-Operate-Transfer methodology support.
4. OEM/White-label Licensing: Cloud providers and OEMs white-label the Rafay Platform to launch branded AI or GPU cloud services with built-in orchestration, governance, and monetization.

Pricing tiers

Go-to-market motion3 records

Distribution channels5 records

Marketing channels7 records

Core offering

Rafay Systems sells the Rafay Platform, an infrastructure orchestration and workflow automation solution that turns GPU and CPU infrastructure into self-service, multi-tenant, policy-governed AI and cloud-native platforms. The platform delivers Kubernetes lifecycle management, GPU cloud orchestration with NVIDIA MIG partitioning, token-metered AI inference monetization (Token Factory), observability via Atlas AI, and air-gapped sovereign deployment. Customers include enterprises, neoclouds, telcos, sovereign AI clouds, and cloud service providers that buy recurring node/cluster subscriptions, usage-based token billing, and OEM/white-label licenses.

Product overview

Rafay Systems offers the Rafay Platform, a unified infrastructure orchestration and workflow automation platform that transforms GPU and CPU infrastructure into self-service, governed AI and cloud-native platforms. The platform operates as a platform-plus-modules architecture where the core Rafay Platform provides foundational orchestration, multi-tenancy, and governance capabilities. Key product modules include: AI Factory for turning infrastructure into self-service AI platforms; Token Factory for token-metered AI service monetization; GPU Cloud Orchestration for multi-tenant GPU service delivery; Kubernetes Management for cluster lifecycle operations; and Kubernetes-as-a-Service, SLURM-as-a-Service, VMaaS, and Bare Metal GPUs-as-a-Service for various compute delivery models. Additional modules include ML Workbenches, Inference-as-a-Service, Self-Service Compute, Observability, Cloud Cost Optimization, Private Cloud Suite, Public Cloud Suite, Standardization Suite, Serverless Inference, and Token Delivery Network. Together these products enable enterprises, neoclouds, sovereign AI clouds, telcos, and cloud providers to move from managing raw GPU infrastructure to delivering monetizable AI services.

Differentiator

Problem solved

Functional benefit

Products and services

• Rafay Platform Core infrastructure orchestration and workflow automation platform that transforms GPU and CPU infrastructure into self-service, multi-tenant, governed AI and cloud-native platforms across public clouds, private data centers, sovereign environments, and edge deployments.
• Token Factory Monetization and metering layer that enables AI factory operators and neoclouds to offer token-based access to AI models via API endpoints with quota management, pricing, and policy enforcement. Validated with OpenClaw and NVIDIA NemoClaw agentic frameworks.
• AI Factory Platform capability that combines GPU orchestration, multi-tenancy, self-service access, policy enforcement, and service catalogs to transform infrastructure into a self-service, governed AI platform delivering applications and AI services at scale.
• Kubernetes-as-a-Service Delivers Kubernetes clusters as elastic, multi-tenant cloud services launched with a single click through self-service portals with approval trails and built-in governance.
• SLURM-as-a-Service Delivers SLURM clusters as elastic, multi-tenant HPC services enabling enterprises to offer HPC infrastructure through self-service consumption for high-performance computing workloads.
• Virtual Machines-as-a-Service (VMaaS) Delivers GPU- or CPU-based virtual machines as consumption-based services through the platform, supporting live migration and snapshot capabilities with isolated compute.
• Bare Metal GPUs-as-a-Service (BMaaS) Enables elastic provisioning of bare metal GPU servers with dedicated GPU access and no virtualization overhead, optimized for large-scale AI training workloads.
• GPU Cloud Orchestration Enables launching multi-tenant GPU clouds that deliver self-service consumption of enterprise-grade AI use cases, supporting GPU pooling, partitioning via NVIDIA MIG and time-slicing, and elastic provisioning of bare metal GPU servers.
• Kubernetes Management End-to-end Kubernetes operations and lifecycle management across public cloud, private cloud, bare metal, and edge environments including cluster provisioning, upgrades, policy enforcement, blueprinting, fleet operations, and add-on management.
• Inference-as-a-Service Cloud-based approach to deploy and manage ML models for real-time predictions without on-premise infrastructure, offering scalability, automated workflows, and GPU management for inference endpoints.
• Token Delivery Network Deploys, governs, meters, and monetizes distributed AI inference services across programmable edge environments, serving AI responses from the best available endpoint based on proximity, performance, policy, sovereignty, capacity, and cost.

Quantifiable outcome

• 20-25% developer time regained
• +8 more outcomes

Named customers22 records

Segments5 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration47 records

AI capability7 records

Feature7 records

Partnerships

17 partnerships are on record, tiered core ecosystem and flagship.

• Ciscocore ecosystemStrategic or Co-development Partner · 2 June 2026Strategic partnership in Elevate ecosystem. Cisco AI PODs transformed into self-service GPU cloud platforms using Rafay. Joint reference architecture and whitepaper on AI POD to GPU cloud transformation.
• UnisysflagshipStrategic or Co-development Partner · 20 May 2026Partnership to provide enterprises with unified platform for scaling and managing AI workloads across public, private, and hybrid cloud environments. Combines Unisys AI expertise with Rafay's infrastructure orchestration for hybrid cloud orchestration, cost transparency, and automated governance for AI agents.
• Dell TechnologiesflagshipStrategic or Co-development Partner · 19 May 2026Deepened collaboration for AI orchestration. Dell positions Rafay as preferred GPU PaaS partner for PowerEdge server customers. Joint go-to-market for AI infrastructure solutions combining Dell hardware with Rafay orchestration.
• NVIDIAcore ecosystemTechnology or Integration · 13 May 2026NVIDIA AI Cloud-Ready validated ISV. Platform supports NVIDIA GB200/GB300 NVL72, HGX B100/B200/B300, RTX PRO 6000 servers. Integration with NVIDIA NIM, NeMo, Base Command Manager, and Unified Fabric Manager. Platform powers NVIDIA-validated reference deployments across the ecosystem.
• AI Green Data Centers (AI-GDC)flagshipStrategic or Co-development Partner · 6 May 2026Sovereign AI infrastructure partnership for Latin America. AI-GDC building Latin America's first high-performance computing and AI platform with Rafay orchestration. Flagship green data center campus in Monterrey targeting 100MW by 2030.
• Argentum AIflagshipStrategic or Co-development Partner · 8 April 2026Software orchestration partnership enabling Argentum to provision and manage separate software environments for individual AI infrastructure customers through single control plane. Positions Argentum to build own neocloud offering and scale from infrastructure broker to full-scale cloud service provider.
• DataDirect Networks (DDN)core ecosystemTechnology or Integration · 19 March 2026Collaboration to help enterprises and neoclouds operationalize large-scale AI infrastructure. DDN's high-performance data intelligence platforms combined with Rafay's infrastructure operations and lifecycle management for streamlined GPU-powered AI workload deployment.
• OoredooflagshipStrategic or Co-development Partner · 15 December 2025Partnership to launch AI-powered Platform-as-a-Service in Qatar leveraging NVIDIA technology. Supports local enterprises in banking, healthcare, and energy to build and deploy AI applications within Qatar's sovereign AI cloud.
• Moro HubflagshipStrategic or Co-development Partner · 23 October 2025Strategic partnership at GITEX 2025 to deliver sovereign GPU-powered PaaS for AI applications in UAE. On-demand access to GPU infrastructure, generative AI services, and AI applications with integrated governance and compliance features.
• Cassava TechnologiesflagshipStrategic or Co-development PartnerDeploying Africa's first NVIDIA-powered AI factories using Rafay Platform. Joint go-to-market for sovereign AI infrastructure across Africa. Deployments spanning Africa with additional partners.
• Firmus TechnologiesflagshipStrategic or Co-development PartnerGreen-energy-powered Australian AI Cloud powered by Rafay's PaaS capabilities. Firmus reports 2x higher GPU monetization using Rafay platform.
• MiTAC Computing Technologycore ecosystemTechnology or IntegrationMiTAC showcasing turnkey AI infrastructure solutions at COMPUTEX 2026 with Rafay for infrastructure management orchestration alongside DDN for AI data platform solutions. 4U AI Powerhouse server featuring Rafay integration.
• Aviz Networkscore ecosystemTechnology or IntegrationPartnership to deliver full-stack GPU cloud orchestration with AI-ready networking. Combined ONES/Spectrum-X networking with Rafay orchestration for AI factory deployments.
• Fortanixcore ecosystemTechnology or IntegrationIntegration with Fortanix Confidential AI platform for automated self-service provisioning in Confidential Computing environments. Joint solution for secure AI deployments using encryption-in-use and trusted execution environments.
• Accenturecore ecosystemImplementation/ SI/ Consulting PartnerStrategic alignment for enterprise and sovereign AI deployments. Joint whitepaper 'Building AI Value Within Borders' published. Accenture provides regional delivery, professional services, and implementation support.
• Deloittecore ecosystemImplementation/ SI/ Consulting PartnerDeloitte partnership for AI infrastructure deployments. Deloitte badge displayed on Rafay website indicating collaboration on enterprise AI initiatives.
• Penguin Solutionscore ecosystemOEM/ Whitelabel/ Licensing PartnerInfrastructure partner helping organizations design, deploy, and operationalize AI infrastructure at scale using Rafay platform.

Scale indicators10 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Run:AI: Run:AI is a GPU orchestration and resource optimization platform for AI workloads — the most direct competitor to Rafay's GPU virtualization and MIG partitioning capabilities. Acquired by NVIDIA in 2024, making it a vertically integrated incumbent within Rafay's primary ecosystem.
• Domino Data Lab: Enterprise MLOps platform that orchestrates AI/ML workloads, model training, and deployment across hybrid infrastructure. Overlaps with Rafay's ML workbenches, Kubeflow-based pipelines, and self-service AI infrastructure for enterprise data science teams.

Emerging players

• Weights & Biases: MLOps platform for experiment tracking, model management, and ML workflow orchestration. While more focused on the data scientist experience, it overlaps with Rafay's AI infrastructure management, MLflow integration, and model serving capabilities.
• Anyscale: Creator of Ray and provider of Anyscale Platform for distributed compute and AI workloads. Direct overlap with Rafay's Ray-as-a-Service and KubeRay offerings, both targeting the data scientist and ML engineering persona on Kubernetes infrastructure.

Others

• CoreWeave: GPU-focused neocloud that is a Rafay customer and ecosystem partner. CoreWeave represents the neocloud buyer persona Rafay serves — its growth signals demand for Rafay's orchestration platform while also illustrating the competitive scale neoclouds must achieve.
• Lambda: GPU cloud provider offering reserved and on-demand GPU instances for AI training and inference. Represents the neocloud buyer segment Rafay serves with its GPU Cloud Orchestration module, and overlaps with Rafay's Bare Metal GPUs-as-a-Service.
• Paperspace (DigitalOcean): GPU cloud and compute platform acquired by DigitalOcean. Represents the GPU-as-a-Service buyer segment Rafay serves; its integration into DigitalOcean illustrates how neoclouds can be absorbed by larger cloud players, potentially reducing Rafay's addressable white-label partner base.

Broad incumbents

• DataRobot: Enterprise AI platform offering end-to-end ML lifecycle management, model deployment, and MLOps capabilities. Broader and more established than Rafay in the enterprise AI application layer, with overlapping positioning on governed self-service AI infrastructure.
• SUSE Rancher: Open-source Kubernetes management platform acquired by SUSE. Direct overlap with Rafay's Kubernetes Management and Kubernetes-as-a-Service offerings, particularly for multi-cluster lifecycle operations across hybrid and multi-cloud environments.
• Red Hat OpenShift: Enterprise Kubernetes application platform from Red Hat (IBM). Competes broadly in the Kubernetes management and hybrid cloud orchestration space, with significant overlap to Rafay's cluster lifecycle, governance, and self-service capabilities.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/RafaySystemsInc

Compliance5 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles5 records

Funding overview

Funding rounds3 records

Investors8 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Rafay Systems