AI-Pete
Apply for the beta

Self-hosted operational triage

AI-Pete turns scattered alerts
into operational cases.

Mailboxes and existing tools feed one prioritized queue. Every case includes context, chat, and actions protected by approval.

Apply for the beta Explore the scenario

We do not replace your tools. We make them work together.

Demo scenario

db-prod-01

Correlated
Correlated sourcesWazuhMailboxOpenSearch
You

What is happening?

AI
AI-Pete

I connected three signals on the same host: an anomalous login from Wazuh, a SQL error in the operations mailbox, and an application error spike in OpenSearch. The case needs review before any restart.

Proposed actionApproval required
Command
systemctl restart mysql
Host
db-prod-01
Risk
Medium 路 database service
Active casesOrdered by priority
Self-hosted
Data stays in your DC
Least privilege
Mandatory approval
Audit trail

One case.
All the evidence.

AI-Pete follows the sequence that currently spans five consoles: detect, connect, explain, propose. The operator keeps the final decision.

  1. 01
    Wazuh

    Detects the first signal

    A login outside baseline opens the case on db-prod-01.

  2. 02
    Mailbox

    Confirms the context

    The operations mailbox reports a SQL error on the same host.

  3. 03
    OpenSearch

    Finds the recurrence

    Application errors increase in the same time window.

  4. 04
    AI-Pete

    Builds the case

    Evidence, impact, and plan arrive in the contextual chat.

  5. 05
    Operator

    Approves or rejects

    No mutating action runs without a human decision.

Inside your perimeter.
Above your tools.

AI-Pete installs in the customer datacenter or the MSP infrastructure. It queries existing systems; infrastructure data does not leave the perimeter.

  • Data localityConnectors and audit trail remain in the customer environment.
  • Limited permissionsEvery connector uses explicit scopes and least privilege.
  • Tracked actionsProposal, approval, and outcome share one audit trail.
Customer datacenter
WazuhGreenboneUyuniProxmoxOpenSearch
Orchestration AI-Pete Correlate 路 warn 路 propose
IT operator Decides and authorizes

Five connectors.
One context.

Planned MVP Core scope. Every connector contributes to the same operational case without replacing the source product.

Planned MVP Core
Planned capabilities for MVP Core connectors
Connector Domain Reads Correlates Proposes Acts Status
Greenbone Vulnerability management With approval Planned
Uyuni Patch management With approval Planned
Wazuh SIEM / security events Read-only Planned
Proxmox Virtual infrastructure With approval Planned
OpenSearch Logs / search Read-only Planned

This table describes the planned MVP scope, not current connector availability.

Private beta

Validate AI-Pete
on your real stack.

We select IT managers and MSPs to validate integration, triage quality, and the approval gate in real environments.

ProfileSME with an internal IT team or an MSP.
StackAt least one tool in the planned MVP Core scope.
ProcessEnvironment review, pilot scenario, and operational feedback.

Submit your application

Enter your business email. We will contact you if your profile fits the beta.

We use your email only to review and manage your beta application.