FAQ: Data Privacy and AI Safety
1. Which AI engine does the app use, and where does it send data?
The app uses top-tier, industry-standard AI engines (such as models from OpenAI, Anthropic, or Google) through our secure enterprise platform partner, Base44.
No direct API exposure: Our app does not store or manage direct third-party AI accounts or API keys. Everything is routed through a single, secure enterprise integration.
Automatic selection: The system automatically selects the most capable and efficient AI model for each request to generate your communications.
2. Does the AI "learn" or "train" on our firm's confidential data?
No. Absolutely not.
Zero training: Neither the app nor the AI providers train their systems on your firm's communications, client details, or uploaded content.
No long-term memory: The AI does not "remember" your firm's data after generating an email. Every request is treated as an isolated, one-time interaction. Once the output is delivered, the session ends and no data is retained by the AI model.
3. How does the app use our firm's content to match our voice?
Instead of retraining an AI model, the app uses a technique called in-context styling:
When you generate a message, the system simply attaches a few short examples of your existing firm templates or past communications to the request as a temporary style reference.
The AI reads these examples in real time to mimic your firm's tone, voice, and formatting, generates the requested draft, and immediately forgets the reference material.
4. How is our firm's data secured and protected from public exposure?
Private Database Storage: Your firm's content and templates are stored securely inside our private database. They are never uploaded to public servers, given public links, or indexed by search engines.
Strict Role-Based Access: Built-in security ensures that your data is strictly compartmentalized. Only authenticated, authorized users within your firm can access, view, or manage your firm's content.
In-Transit Encryption: When a draft is requested, the necessary style examples are sent over an encrypted channel to the AI provider solely for the duration of that single request. It is processed in memory to draft the response and is never published or made available to outside parties.
Summary for Your Peace of Mind
Your firm's proprietary information is never public, never used to train AI models, and never shared with other firms. The AI acts strictly as a temporary assistant that reads your style guidelines on the spot, writes the draft, and immediately clears its workspace.