Sentilai

Sentilai is an AI control plane for software teams: governance, visibility and policy enforcement over how AI coding assistants and agents are used.

• Gateway — a provider-compatible proxy for Anthropic, OpenAI, Azure OpenAI, Gemini, Mistral and OpenAI-compatible providers; per-tenant policies, token accounting and audit on every call
• MCP governance — per-tool policies, install vetting, protocol-posture checks and real-time enforcement for Model Context Protocol servers
• Risk engine — synchronous detection of secrets, PII and risky agent-autonomy patterns, with configurable block/warn actions and real-time Slack/Teams/webhook alerts
• Endpoint governance — signed, auto-updating agents for macOS and Windows with device identity, one-click "govern all" and one-click AI offboarding
• Audit & SIEM — an append-only audit trail customers can pull into their own SIEM via API

Security posture: passkey-first (WebAuthn) authentication, per-tenant identity isolation, AES-256-GCM-encrypted provider keys with versioned key rotation, client-side-encrypted Object-Locked backups with tested restores, EU/US regional data planes with no cross-region data flow, and prompt-content capture that is opt-in and off by default.

Privacy Policy, DPA & sub-processors · Docs · Pricing