Who owns the data and the twin?
You do. Everything you provide (your documents, your interviews, your training material) stays your property, and the twin built from it is yours. If you ever end your engagement, your data comes back to you and is deleted from our systems. We don't use it to train other twins or improve anyone else's models.
How is my voice protected from misuse?
Your voice is an asset and a deepfake risk, so we treat it like one. The platform uses voice verification so your twin only runs in contexts you've authorized, and we're building voice fingerprinting and audio verification to catch unauthorized use. When you're putting a real person's voice into an AI, that protection isn't optional.
Can the twin say something I didn't sanction?
You set the boundaries, and it stays inside them. You decide the topics it can speak to and the ones it won't touch. Every answer traces back to a specific source in your knowledge base, so if something's off, you can see exactly why and correct it. When a question falls outside what the twin knows, it says so in your voice rather than inventing. No black box, and you can see the receipts.
Is it secure enough for serious use?
Yes. We use enterprise-grade infrastructure with encrypted storage and transmission, role-based access, full audit trails, and data isolation between twins, and we're pursuing SOC-2 for enterprise deployments. Your conversations are yours and your team's to see, never shared with anyone else or used outside your engagement.
What if I want to stop?
You can end the engagement anytime. Your data is returned and deleted, with no long-term lock-in. The twin is yours to walk away from, which is part of how we keep your trust.