Disclaimer: SereneAI is a self-guided emotional reflection journal for general knowledge. It is not licensed medical advice; consult a healthcare professional for health concerns.

Hey HN! I built SereneAI after hitting a severe emotional breaking point. I found myself hysterically crying, feeling completely alone, and unable to afford traditional therapy prices. I realized people don't need expensive corporate paywalls when they are drowning—they need an immediate, private space to vent.

To guarantee total data security, I engineered SereneAI to be completely database-free and serverless. Your text and metrics never live on an external server; all context arrays and 7-day mood tracker metrics reside strictly within the user's localized browser memory (localStorage).

Features Implemented: • CBT "Vent and Validate" Prompts: Forces the AI to practice active, reflective listening first, instead of barking rigid, robotic lifestyle advice. • Native TTS Integration: Programmatically filters browser native voice arrays (window.speechSynthesis) to run calming Australian and British English vocal profiles at a slower, therapeutic 0.85x speed. • Interactive Self-Care Toolkit: Features an animated box breathing guide, a cognitive distortion reframer form, and localized journal feeds. • Client-Side Safety Filters: Hardcoded regex keyword check pipelines instantly block and route urgent distress phrases to human resources.

It is completely free to use. I am building this in public as a bootstrapped solo creator and logging updates over at our new community hub https://www.reddit.com/r/Serene_AI/.

I would love to hear your thoughts on the UI layout, the text-to-speech pacing, or how to further optimize client-side context tracking!