01
Platform owner, not reseller
Your brand. Your pricing. Your customer relationship. You own the platform and the margin - not a hyperscaler's distribution channel.
Ready to get started?
Deploy sovereign AI on your infrastructure - in weeks, not months.
AI Cloud · For service providers
White-label the full Katonic platform as your own sovereign AI service. Provision enterprise customers in minutes. Per-org isolation. Per-org P&L. Your brand on every screen.
Reselling a hyperscaler makes you a distribution channel. Katonic makes you a platform owner.
You have the infrastructure. You have the enterprise relationships. You have the data center footprint. What you don't have is a full-stack AI agent platform with sovereign deployment, white-label branding, and per-customer isolation.
Katonic AI Cloud ships as a literal product: same codebase as AI Factory, with a Distributor Console for org lifecycle, GPU allocation, per-org billing, and silo-isolated infrastructure per enterprise customer.
01
Your brand. Your pricing. Your customer relationship. You own the platform and the margin - not a hyperscaler's distribution channel.
02
Dedicated databases, namespaces, Keycloak realm, and GPU queues per enterprise customer. No cross-org data path. Hardware-enforced isolation.
03
Per-org P&L with cost vs. revenue tracking. OpenCost infrastructure cost. GPU pricing you control. Margin analysis per customer.
04
Distributor Console: create org, allocate GPUs, set capabilities, configure branding. Customer logs in to their white-labeled Workroom immediately.
The Distributor Console is the telco operator's view of the platform. Create organizations, allocate resources, set pricing, and track profitability across all your enterprise customers.
Org lifecycle management
Create, suspend, activate, and terminate organizations. View status, quotas, active sessions. Each org gets a fully isolated environment from the moment you click create.
GPU allocation with pricing
Allocate GPU slices with per-slice pricing you control. GPU pools with time-based expiry. Full allocation history for billing reconciliation.
Cross-org user management
Create users, assign roles, manage permissions across all organizations. 8-level RBAC. Per-org capability restrictions below platform license ceiling.
§ Organizations
Production tenants
The Distributor Console lives at /distributor for super-admins. Backed by the tenant-manager service (30 endpoints). See every tenant, drill into per-org config, watch MRR roll up. The chrome below is what your operations team uses to run the AI cloud.
Distributor
Raj Tanaka
TelcoX · super admin
Operations
Configuration
Compliance
Tenants
5 organizations · 115M+ users · $533k MRR · backed by tenant-manager (30 endpoints)
Active tenants
5
+1 this quarter
Total users
115.0M
+8% MoM
GPU pool
41/48
85% utilised
Recurring revenue
$533k
+11% MoM
| Tenant | Custom domain | Users | Agents | GPU | Environment | MRR | Status |
|---|---|---|---|---|---|---|---|
P Pilipinas AI kt-pilipinas · since 2024-Q3 | ai.pilipinas.gov.ph | 115M | 142 | 32× H100 | SOVEREIGN | $480.0k+12% | ACTIVE |
A Acme Bank kt-acme-bank · since 2024-Q2 | ai.acmebank.com | 1,200 | 18 | 4× H100 | PRODUCTION | $24.5k+8% | ACTIVE |
M Metro Hospital kt-metro-health · since 2025-Q1 | agent.metrohealth.org | 450 | 8 | 2× A100 | PRODUCTION | $12.2k+4% | ACTIVE |
C City Gov kt-city-gov · since 2024-Q4 | ai.city.gov.local | 800 | 12 | 2× H100 | AIR-GAPPED | $18.8k | ACTIVE |
R RetailCo kt-retailco · since 2026-Q1 | (default subdomain) | 200 | 4 | 1× A100 | TRIAL | - | TRIAL |
Click any tenant to drill into per-org admin (SSO, custom domain, environments, features)
/distributor in your sandbox today, backed by the real tenant-manager service (30 endpoints) - SSO per org, custom domains, environments, feature flags, and cross-tenant audit. Each tenant is a fully isolated namespace.Per-org isolation architecture
● All isolatedDedicated namespace · databases · SSO · queue per org · no shared data path
Not namespace-level. Not schema-level. Hardware-enforced silo isolation. Every enterprise customer gets dedicated databases, K8s namespace, Keycloak realm, and GPU queues. No cross-org data path exists.
Dedicated databases per org
PostgreSQL, MongoDB, Milvus, Redis, and MinIO instances provisioned automatically by K8s operators. No shared tables or connections.
Per-org SSO realm
Dedicated Keycloak realm with SAML/OIDC federation. Customer's admin manages their own users. No other realms visible.
Network-level isolation
K8s NetworkPolicy blocks cross-namespace traffic. Istio mTLS encrypts service-to-service. No pod can reach another org's pods.
Every organization sees your brand, not Katonic's. You set the pricing. The platform computes costs. The margin is yours.
White-label per organization
Organization name, colors, logo, assistant persona, welcome prompts - all configurable. Custom domains (CNAME + auto-TLS) as premium.
You set the pricing
GPU price per hour per slice. LLM token markup. Storage, CPU, memory pricing. Platform tracks consumption - you set unit economics.
Per-org P&L in real time
Infrastructure cost (OpenCost) vs. revenue = margin per org. Monthly trends. Billing export for your invoicing system.
Infra cost
$7,400
OpenCost
Revenue
$9,768
Your pricing
Margin
$2,368
32% · ▲ healthy
Per-organization breakdown
Philippines · ePLDT · Launched 2025
ePLDT white-labeled Katonic to launch a national sovereign AI platform serving 115 million people. Banks and government agencies as first tenants. Full platform: Workroom for employees, Studio for developers, Control Room for IT. Zero data egress from Philippine infrastructure.
This is the model: a service provider white-labels Katonic, deploys on sovereign infrastructure, and launches an AI business serving enterprise customers across regulated industries.
90 days from contract signing to production. The fastest sovereign AI deployment at national scale.
“Katonic let us launch AI services to our enterprise clients without building the platform ourselves. White-label, multi-tenant, usage-based billing. Exactly what we needed. We were live in 90 days.
See the Distributor Console, silo isolation, and per-org P&L in a live 30-minute demo.
Learn how ePLDT launched Pilipinas AI in 90 days.