Telecom, managed services, technology
White-label AI service for MSPs and telcos
How can a managed-service provider sell private AI to its customers on its own infrastructure?
Katonic lets a managed-service provider or telecom operator sell each of its customers a private, branded AI platform that runs on the provider's own infrastructure. Each customer is isolated and metered separately. The provider gains a sovereign AI product to sell and a way to earn revenue from its GPUs.
At a glance
| Industry | Telecom, managed services, technology |
|---|---|
| Who it is for | MSP and telecom CEO, Chief Product Officer, Head of B2B products |
| Data that stays in place | Each customer's models, agents, knowledge and usage, inside the provider's infrastructure |
| Runs air-gapped | Depends on the provider's service design |
| Commercial model | Provider bills its own customers from metered usage |
| Infrastructure | HPE Private Cloud AI, or HPE GPU servers in an HPE AI Factory design |
The problem
Managed-service providers and telecom operators own GPU capacity and customer relationships but have no AI product to sell.
Each of their customers wants its own private AI platform, with its own models, agents and knowledge, and the provider cannot build one from scratch for each.
How it works
- The provider installs the Katonic platform on its own infrastructure.
- It creates an environment for each customer, under the provider's own brand.
- Each customer gets its own models, agents and knowledge, isolated from every other customer.
- Usage is metered per customer.
- The provider bills its customers and manages them from the distributor console.
On HPE infrastructure
On HPE Private Cloud AI
HPE Private Cloud AI is HPE's turnkey private AI system, co-engineered with NVIDIA. Katonic AI is a partner in the HPE Unleash AI program. On Private Cloud AI, Katonic connects to the models the system serves and adds the agent platform on top. A single organisation can run the same platform on Private Cloud AI for its own teams. A provider that hosts many customers is usually better served by the HPE GPU server route below.
On HPE GPU servers
Customers who build their own estate on HPE GPU servers, as part of an HPE AI Factory design, get compute, storage and networking from HPE. Katonic supplies the software layer above the hardware: open-weight models served on the customer's GPUs, the AI gateway, the agent platform and governance. The provider builds its AI cloud on HPE GPU servers and uses Katonic as the service layer that customers sign in to.
More on both routes: Katonic AI on HPE.
Where it stands
Live with a telecom AI cloud in Southeast Asia.
Questions and answers
Whose brand does the customer see?
The provider's. The platform is offered under the provider's own name.
Can one customer see another customer's data?
No. Each customer has its own isolated environment with its own models, agents and knowledge.
How does the provider bill?
Usage is metered per customer, and the provider bills from those records under its own pricing.
Where does the service run?
On the provider's own infrastructure, in the provider's own country or region.