AI-powered supportive conversations
Daniela provides calm, nonjudgmental conversations. It asks clarifying questions, reflects its understanding of the situation, suggests one coping technique at a time and helps the user identify a small, practical next step.
The application does not impersonate a human, offer medical or psychological diagnoses, or replace professional advice.
Personal memory and continuity
Daniela retains relevant context from earlier conversations so users do not have to repeat their stories each time.
A dedicated interface allows users to review, correct and remove saved information. Personal memory is therefore designed to support continuity while remaining manageable by the user.
Rehearsing difficult conversations
Users can describe someone from their lives and practice a difficult conversation with an AI representation of that person.
After the rehearsal, Daniela provides feedback and allows useful phrases to be saved for reference before a real conversation.
When signs of violence or threats are detected, the system avoids encouraging potentially dangerous confrontation and directs the user toward assistance instead.
Risk assessment and crisis assistance
Every message undergoes a risk check before the ordinary AI response is generated.
The process combines rules for identifying explicit risk-related expressions in Hebrew and Russian with a separate model that evaluates conversational context.
At moderate risk, Daniela may ask about the user's safety and present support contacts. At high risk, the ordinary response is interrupted and a crisis assistance screen appears.
The screen provides country-specific assistance contacts, including resources for Israel and Russia, with one-tap calling. The contact information remains available without an internet connection.
A permanent “I don't feel safe” control also allows users to access assistance without waiting for automatic detection.
Risk assessment is an additional safety measure, not a guarantee that every emergency will be recognized.
Privacy and on-device safety
- PIN protection: Controls access to the application.
- Secondary PIN: Opens an empty application view without displaying conversation history.
- Quick exit: Immediately switches to a neutral screen.
- Delete everything: Clears information stored on the device through a single action.
- On-device encryption: Personal information stored on the device is encrypted.
- Selectable app icon and neutral notifications: Additional options intended to reduce unwanted exposure of app usage.
These measures reduce certain privacy risks but cannot guarantee complete protection in every situation.
Additional capabilities
Screenshot-based conversation analysis: Users can submit screenshots of conversations for analysis to help explore communication patterns and difficult interactions.
AI-initiated messages: The application can initiate communication rather than responding only when the user sends a message.
Practical information and exercises: Content covering subjects such as divorce and workplace rights, intended to support understanding and preparation without replacing legal or professional advice.
Horoscope: An additional content feature that is separate from clinical assessment and crisis risk detection.