Adfinia. ← Back to adfinia.com
Feature · Enterprise

Sovereign AI. On your GPU, inside your jurisdiction.

AI inference for Adfinia, pinned to a local LLM you control. Customer data never leaves your environment. No outbound call to a US-hosted model API. Toggleable per tenant; one switch, full audit trail.

AvailabilityEnterprise tier only
TogglePer-tenant · admin-gated
Cooldown30 days both directions
AuditVerifiable bypass of the cloud AI provider
This is a working draft, pending external legal review. The feature description below reflects our current product design. Hosting topology, supported model list, and regulatory positioning are being finalised with external counsel and may change before commercial publication. For an architectural deep-dive, write to support@adfinia.com with subject "Sovereign AI".

Why sovereign matters

Generic marketing platforms route AI prompts to a US-hosted model. Your prompts include customer data. That's a lot of trust to outsource to someone else's transit network.

01

Data residency, end-to-end

Prompts, completions, embeddings, and any retrieved context never leave the geographic boundary you select — including the AI hop. The model runs on your GPU, in your network.

02

No third-party processor

No SaaS LLM provider sub-processor for sovereign-flagged tenants. Your DPA doesn't need to list a model vendor; your audit doesn't need to ask one for a SOC 2.

03

Procurement-friendly

Built for procurement, defence, healthcare, and regulated public-sector tenants that can't accept outbound model-API traffic. We pin the AI router, then publish proof.

How the routing works

Every AI feature in Adfinia goes through one internal service — the AI Gateway. Sovereign mode is a flag on the tenant that switches that gateway's upstream from the cloud model API to a local inference service.

Cloud mode (default)
Adfinia API AI Gateway cloud-ai-provider.api completion
Sovereign mode (tenant flag)
Adfinia API AI Gateway local-llm.svc.cluster.local completion
Verifiable in the admin console: every AI invocation logs the destination URL, the model identifier, and the bytes transferred. The audit query SELECT COUNT(*) FROM ai_invocations WHERE destination LIKE '%cloud-ai-provider%' AND tenant_id = '…' must return zero for a sovereign tenant.

The 30-day cooldown

Sovereign isn't a feature flag you flip in a meeting and flip back the next morning. It's a posture change with operational and contractual consequences — so we put a 30-day cooldown on both directions of the switch.

Cloud → Sovereign

30 days to provision

Once you flip on Sovereign, the next AI invocation routes to your local LLM. The flag itself is locked for 30 days — you cannot switch back to cloud sooner. We use the window to confirm your model is stable, your performance is baselined, and your audit trail shows zero cloud egress.

Sovereign → Cloud

30 days to wind down

If you want to switch back to cloud (typically because you've outgrown your on-prem capacity), the flag is locked for 30 days the other way. The window gives your security team time to update DPAs, re-list sub-processors, and notify regulators where required.

The cooldown is enforced at the platform layer — not a soft setting. The admin console shows the unlock date; a super-admin can force-unlock with a typed reason that hits the audit log and a #admin-audit Slack notification.

Jurisdictional pinning

Sovereign tenants pin both data residency and AI inference to one jurisdiction. The current options:

UAE

Abu Dhabi · Dubai

In-country inference for PDPL-regulated workloads and government tenants. Mistral, Falcon, and licensed Llama variants supported.

India

Mumbai · Hyderabad

In-country inference for DPDP-regulated workloads. Aligns with MEITY guidance on cross-border data flow.

EU

Frankfurt · Stockholm

EEA-resident inference for GDPR-restricted workloads. Compatible with EU AI Act transparency obligations.

KSA

Riyadh

In-country inference for SDAIA-regulated workloads. Available on contract; provisioned via local partner.

Kenya

Nairobi

In-country inference for ODPC-aligned workloads. Roadmap; talk to us if this is a procurement requirement.

On-prem

Customer cluster

Deploy the inference service inside your own Kubernetes cluster. We bring the model wrapper; you bring the GPU.

Switching jurisdictions inside Sovereign mode is itself subject to the 30-day cooldown. See the /legal/regions page for the underlying data-plane residency table.

What you give up — honestly

Sovereign is a posture choice, not a free upgrade. Some features that depend on cloud-hosted models behave differently:

We won't silently fall back to a cloud model. The UI tells you, the audit log records it, and your DPO can verify it.

Talk to us

Sovereign is an Enterprise feature and a 4-week onboarding — there is no self-serve path. The form below routes to our solutions engineering team; we usually reply within one business day.

Request a Sovereign briefing

Tell us about your workload — model preferences, jurisdiction, GPU footprint, audit constraints. We'll come back with a 45-minute architecture call and a quote.

Submitting this form is consent for Adfinia to reply by email. We do not pass your details to any third party. See /privacy.