Investors

Building the control layer for the AI economy

AI adoption is creating a new enterprise control layer: the place where an organisation discovers its AI, decides what is allowed, enforces that decision and proves it. MERIDVAR is building that layer for the buyers who must run it themselves.

  • Category: Enterprise AI Control Plane
  • Buyer: the CISO of a regulated enterprise
  • Delivery: customer-operated appliance
  • Stage: early-stage company, product in beta

The problem

AI use has outrun AI governance

Security, privacy and risk teams are asked to govern AI, yet the five questions they must answer rarely live in one place: what is in use, how risky it is, who approved it, what is allowed through, and whether any of it can be proven.

13%of organisations reported a breach of AI models or applications.
97%of those organisations lacked proper AI access controls.
USD 670,000was added to the average cost of a breach where shadow AI use was high.

Source: IBM, Cost of a Data Breach Report 2025.

Why now

Governance has become a budget line

A named category

Analysts now track AI governance platforms as a distinct software category, separate from AI security and from GRC.

Regulation in force

EU AI Act transparency obligations have applied since 2 August 2026. Prompts that carry personal data already fall within the GDPR.

Consolidation under way

Since 2025, several of the largest security vendors have acquired AI security and AI gateway companies. The category is consolidating into platforms.

Positioning

The open question is who connects the decision to the traffic

Governance tools record decisions. Security tools inspect traffic. Regulated buyers need both in one record, on infrastructure that their own vendor risk process can approve. That is the gap MERIDVAR is built for.

A specific buyer
The CISO of a regulated enterprise, with privacy, risk and compliance as co-sponsors.
A specific constraint
No vendor cloud and no new data processor in the AI data path.
A shipped product
Five connected stages in one appliance, in evaluation on customer infrastructure.
A stated boundary
We do not target buyers who are comfortable with cloud processing. An incumbent will usually serve them well.

Product

One appliance and one record, operated by the customer

MERIDVAR reference architectureCallers on the customer network reach AI providers through the MERIDVAR appliance, which runs five connected stages: discover, assess, approve, control and prove. Security, privacy, risk, finance and audit teams read one shared record. Nothing leaves the customer network for MERIDVAR.Your networkOutside your networkPeopleA token for each personAI applicationsOpenAI-compatible APIAI agentsSigned identitiesMCP serversDeclared inventoryUnmanaged trafficPAC file and proxy captureMERIDVAR applianceOne organisation, one appliance, and no control plane in our cloud01DiscoverInventory and shadow AI02AssessQuestionnaires and risk rules03ApproveReviews and approvals04ControlAdmission gateway05ProveEvidence log and audit packageModel providersAnthropic, OpenAI, Azure OpenAI, xAISelf-hosted modelsYour own endpointSaaS AI servicesSeen through uploaded logsAdmitted requests onlyNo telemetry to MERIDVARRuns without internet accessOne shared recordSecurityPrivacyRisk and complianceFinanceAuditEvery team that answers for AI reads the same appliance: visibility, policy, control, evidence and spend.
Controls apply to traffic routed through the gateway. Visibility of other traffic depends on uploaded logs, endpoint agents and declarations.

Strategy

A record that becomes more valuable with use

Several budget owners

Security buys. Privacy, risk and compliance co-sponsor. Finance reads the ledger. One deployment answers to all of them.

Modular expansion

Controls are licensed as modules on a subscription. A customer starts with a subset and adds modules on the same appliance, with no redeployment.

An embedded system of record

Once approvals, reviews and their history live in MERIDVAR, replacing it means migrating the evidence. That switching cost has to be earned with customers. It does not exist before them.

A widening surface

Agents and MCP servers add new subjects to govern. The same inventory, rules and evidence log extend to each of them.

Where we stand

An early-stage company, described plainly

MERIDVAR is an early-stage company. The product ships and is being tested on customer infrastructure. We describe the company with the same discipline that the product applies to evidence: no invented traction, no borrowed logos and no unsupported market arithmetic. We do not claim a moat. We intend to earn one.

Further material is shared in conversation, under a confidentiality agreement.