When you connect an AI to your inbox, you're trusting it with your most sensitive information. We don't take that trust lightly.
No Training on Your Data
We use state-of-the-art models from Anthropic (Claude), but we have strict agreements in place: your data is never used to train their models. Your business remains your business.
Encryption at Rest and in Transit
Every piece of data, from your OAuth tokens to the cached summaries, is encrypted using AES-256-GCM. We use industry-standard security protocols to ensure that even if someone gained access to our servers, your data would remain unreadable.
Transparency by Design
We publish our system prompts at /transparency. You can see exactly what instructions we give the AI when it reads your emails. No hidden agendas, no "black box" logic.