S2 PRIVATE AI / PRIVATE INFRASTRUCTURE

Your AI. Inside your business.
Ready for the first question. Open to what comes next.

Private AI infrastructure deployed in your environment, connected to selected business tools and supported by the S2 team.

Book a discovery call See what is included
01Client-owned hardwareOne-time purchase; no vendor lock-in.
02Private by designYour deployment boundary is agreed before build.
03Integrated for workA real first use case, not a demo-only box.
04Open next stepsAdd systems and workflows as needs grow.

PRIVATE AI, MADE PRACTICAL

Control over your data. The capabilities of modern AI.

S2 Private AI combines local AI infrastructure, open ecosystem models and integration with business workflows. Together we define what stays in your environment, how data is accessed and which actions the system may take.

This is not a rental of a closed appliance. You purchase and own the hardware; S2 brings assessment, configuration guidance, installation, integration, support and ongoing engineering.

THE FIRST DEPLOYMENT

From infrastructure to a useful first result.

The initial scope is confirmed after discovery and an environment review.

Discover

Understand your data, policies, users and target use case.

Deploy

Prepare the local AI runtime and access for named users.

Connect

Configure one selected channel and an initial automation.

Enable and support

Document the setup, onboard users and agree ongoing care.

  • Local GPU server, approximately 24–48 GB VRAM, subject to confirmed configuration
  • Up to 20 named Open WebUI users
  • One private communication channel
  • One n8n automation and one agreed first use case
  • Configuration, initial training and documentation
  • Hardware remains the client’s property

OPEN, REVIEWED STACK

Components follow your requirements.

This is a proposed direction, not a locked vendor list. Versions, licensing, security and feasibility are confirmed for each project.

vLLM

Model serving for supported language models.

proposed · validate

Open WebUI

Private interface for working with AI.

proposed · validate

Keycloak

Identity and access management.

proposed · validate

PostgreSQL + pgvector

Storage and retrieval for approved context.

proposed · validate

n8n

Approved workflow orchestration; verify licensing.

proposed · validate

MCP gateway

Controlled tool access with separate permissions.

proposed · validate

SECURITY AND BOUNDARIES

Private still needs clear rules.

System access and AI agent actions are permissioned separately. Read access to an ERP does not grant permission to send payments, delete or change records.

Local processing can limit prompts sent to external models, but security depends on configuration, patching, network controls and procedures. External channels such as WhatsApp or Signal may process data through their providers and must be assessed separately.

INVESTMENT

You own the hardware. S2 stays alongside.

AI Care

from €300 / month

Proposed monthly support package for Private AI; final SLA and scope are agreed separately.

Enterprise configuration

Scoped together

More users, redundancy, integrations and stronger security requirements.

Figures are indicative and exclude VAT. Hardware warranty, advanced SLA, additional integrations and licensing are excluded until specified in a proposal.

FAQ

Frequently asked questions

Does all data stay local?
A local model and infrastructure can process data in your environment, but integrations and external channels may involve other providers. We define these flows first.
Who owns the hardware?
The client purchases and owns the device. S2 provides assessment, installation, integration, support and engineering.
Can the AI change business records?
Only when that access is explicitly integrated and approved. Read and write permissions are configured separately.
Are these final prices?
No. They are indicative target prices excluding VAT. A final proposal follows a configuration and scope review.

START WITH YOUR ENVIRONMENT

First, let’s see if Private AI fits your use case.

The S2 team reviews your security requirements, existing systems and first business use case with you.

Request a conversation
Chat

AI Assistant