We build governed AI agents that turn live public-web evidence into evidence-ranked, action-bounded, WAL-verifiable risk intelligence.