33 Logo

Description

Built AuthRail, an AI governance platform that helps organisations control agent behaviour and manage compliance. Built runtime controls that enforce policies before agents act, alongside compliance workflows for managing regulatory obligations, controls, and audit evidence, helping teams identify gaps and audit agent activity against organisational policies. Developed Node.js, Go, and Python SDKs, MCP integrations, and NVIDIA OpenShell support to embed governance into agent workflows and make agent activity traceable to the rules and approvals governing it.

Highlights

  • Built AuthRail, an AI governance platform that turns organisational policies into enforceable rules for AI agents, covering which actions they can take, which data they can access, and what they can share.
  • Built a runtime authorization system that intercepts agent tool calls before execution and applies business policies to allow actions, block them, or require human approval.
  • Developed Node.js, Go, and Python SDKs, MCP integrations, and NVIDIA OpenShell support to bring authorization, sensitive-data protection, and execution isolation into agent workflows.
  • Built compliance management workflows connecting regulatory frameworks, organisational policies, controls, and evidence, helping teams track obligations, identify gaps, and manage remediation.
  • Designed versioned Rails and an AI inventory to manage agent permissions, review policy changes, and trace agent activity back to the rules and approvals governing it.