Skip to main content
Altrum governs the AI your organization builds and buys, from the first question about a new AI system to every request it sends to a model. You keep a register of your AI systems. Each one is assessed in a conversation with Altrum’s assistant, which finds the risks that apply and the controls that answer them. The controls switch on guardrails, which check what people send to the AI system and what the model answers. When the AI Gateway arrives, your applications send their model traffic through Altrum, so the guardrails, budgets and access rules apply to every request, and you see what happened.

How Altrum works

The pieces and how they connect: AI systems, assessments, risks, controls, guardrails and the AI Gateway.

Start by role

Reading paths for governance and risk, security, AI builders and administrators.

Quickstart

From your invitation to live guardrails on your first AI system.

AI Gateway

Send your applications’ model traffic through Altrum. Coming soon.

​
What you can do today

  • Invite your colleagues and give each a role: Members and roles.
  • Register the AI systems your organization uses: AI systems.
  • Assess an AI system and review its risks and controls: Assessments.
  • Configure, test and apply guardrails, with suggestions from the assessment: Guardrails.
  • Ask for help from inside the product: Support.

​
What is coming

  • The AI Gateway: one address for your applications and coding agents to reach their model providers, with guardrails, budgets, rate limits, model access, MCP tools and request logs. See AI Gateway.
  • Guardrails for your organization: guardrails that apply to gateway traffic not tied to a registered AI system. See Organization guardrails.
  • Reports and dashboards: governance posture, guardrail activity and cost across your AI systems. See Reports.
Pages for features that are not available yet carry a Coming soon tag in the sidebar.