MCP/SI Lab
ENGet started
THE GOVERNED INTELLIGENCE GATEWAY

Your systems.
Your agents.
Your rules.

Turn internal systems into governed MCP servers. Give intelligence a way to act — and your company the controls to trust it.

Start buildingExplore the platform
Human approval. Scoped tools. Traceable actions.
LIVE ARCHITECTUREMCP / 01
YOUR INFRASTRUCTURE
ERP & CRMBusiness operations
APIs & databasesStructured data
SharePointOrganizational knowledge
MCP GATEWAYGOVERNED
Policy engine
Approval queue
Audit trail
Your security boundary
✳Claude
◉ChatGPT
◈Copilot
⌘Your agents
COMPATIBLE MCP CLIENTS · CONFIGURATION REQUIRED
01 / CONNECT → GOVERN → ACT
ONE PROTOCOL. CONTROLLED ACCESS.
PostgreSQL + SharePoint + CRM & ERP APIs → MCP tools → Governed execution → OAuth for Claude & ChatGPT
FROM CONNECTION TO CONTROL

Make your systems
agent-ready.

A single workspace for the people building AI access and the people responsible for it.

MCP/SI Lab
A Acme workspace
Illustrative workspace
Workspace / Connectors PRODUCT PREVIEW
YOUR INTELLIGENCE INFRASTRUCTURE

Connect once. Put it to work.

Operational
Connected sources03↗
Published tools08↗
MCP servers02↗
Pending approvals02↗
Connected sourcesSAMPLE DATA
ERP APIREST APIConnected
Product catalogJSON datasetConnected
CRM APIREST APIConnected
Control is built into the workflow.

Publish explicit capabilities. Approve writes. Keep a record.

Open workspace
A DELIBERATE WORKFLOW

From siloed systems
to useful intelligence.

Read the documentation
01

Connect your sources

Connect PostgreSQL, SharePoint, Dynamics 365, Salesforce, HubSpot or any approved HTTPS API. Credentials stay in an encrypted per-workspace vault.

02

Compose your tools

Compose each tool on a visual canvas: source, operation, privacy, governance and publication.

03

Govern every action

Mask personal data, require approval for sensitive actions and separate builders from reviewers.

04

Publish your server

Publish an immutable tool snapshot. Agents connect with OAuth 2.1 or a revocable scoped key.

GOVERNANCE IS THE FOUNDATION

Give agents capability.
Keep the authority.

Intelligence should work within your organization’s boundaries. Set those boundaries before the first tool call.

Explore the controls

Explicit access

Agents receive only the tools published to their server. OAuth grants and keys are scoped to one server and stored as hashes.

Human decisions

Write requests enter an approval queue. A different reviewer approves the exact arguments; the agent retrieves the outcome.

Traceable execution

Follow tool execution, approval decisions and configuration changes in a persistent audit log.

An immediate off switch

Pause a server, revoke OAuth grants or rotate its key. Published snapshots stay independent of draft changes.

Roles and privacy

Owner, admin, builder, approver and viewer roles. Personal data in responses is masked before it reaches the model.

PRICING

Priced for production,
not for experiments.

Prices in EUR, excluding VAT. Billed monthly. Implementation quoted separately.

Starter

One team, first governed agents.

€490/ month
  • 3 MCP servers
  • 5 members
  • All standard connectors
  • OAuth and scoped keys
  • Approvals and audit log
  • Personal data masking
Start a pilot

Business

Departments running agents in production.

€1,200/ month
  • Unlimited MCP servers
  • 25 members and all roles
  • Company SSO (SAML)
  • PostgreSQL and enterprise connectors
  • Audit export
  • Priority support
Talk to us

Enterprise

Regulated organizations and dedicated deployments.

From €2,500/ month
  • Dedicated EU deployment
  • Unlimited members
  • Custom connectors
  • SLA and named support
  • AI Act documentation support
  • Security review assistance
Contact sales

Implementation and integration: from €5,000 per organization, depending on systems and scope.

A FEW THINGS TO KNOW

Built on clarity.

Understand the platform before connecting your systems.

YOUR NEXT LAYER OF INFRASTRUCTURE

Connect intelligence.
Retain control.

Build your workspace