Obstruo.ai

Obstruo.ai is an AI governance and compliance tool that acts as a neutral control plane in front of every AI provider and MCP server, so regulated enterprises can govern what their AI can see, say, and do, and prove it on every call. It is designed as a drop-in proxy: you point your existing SDK at an OpenAI-compatible endpoint and change nothing else in your code, and existing SDKs keep working through one key and one endpoint. Obstruo centralizes API keys, applies per-jurisdiction PII redaction automatically, keeps complete audit logs for every call, and adds built-in safety against prompt injection. Because it is provider-agnostic, it routes across different LLM vendors and MCP endpoints without lock-in. The platform separates two consoles: an engineering view for endpoints, connections, models, redaction, guardrails, and monitoring; and a compliance view covering posture, an exposure map, a policy catalog, an asset and vendor registry, DPIA and ROPA records, data-subject rights, and approvals. It tracks framework readiness against EU AI Act, GDPR (including Articles 30, 32, and 35), SOC 2 Type II, and ISO 27001, keeping controls current from live traffic. A regulatory exposure map flags any agent that reaches data more sensitive than its own classification allows, and an approvals queue lets a compliance officer sign off or send back each agent, MCP server, or raw-audit exemption with the risk and owner attached. Obstruo.ai is EU-resident and positioned for compliance and engineering teams that need centralized keys, automatic redaction, and unified audit for enterprise AI.

Compliance and engineering teams at regulated enterprises use Obstruo.ai to govern, redact, route, and audit every AI call across providers from one neutral control plane.