Hello everyone, I am launching ZultaAI today as a solution for on-device AI productivity. The project was built on a single architectural principle: keeping all user data and AI inference strictly on the local hardware. By utilizing Google’s Gemma 4 E2B model, ZultaAI provides transcription, summarization, and a searchable knowledge base without the need for cloud infrastructure or data-sharing agreements. For those testing the application, please keep in mind: Model Initialization: Core operations require an initial 2.8GB model download to your device. Hardware: Performance is optimized for modern mobile SoCs with 6GB+ of RAM. Encryption: All storage is secured via SQLCipher and biometric locking by default. I am particularly interested in technical feedback regarding the on-device inference speed and the accuracy of the action item extraction. I will be available throughout the day to discuss the implementation details or any specific feature requirements you may have. Thank you for your time. NOTE: App is only available for 6gb+ ram devices. SAVE with 50% launch discount!

ZultaAI: Encrypted AI Notes

Organize thoughts with AI that never leaves your phone.

Organize thoughts with AI that never leaves your phone.