Hey everyone! I'm launching ReflectAI on Product Hunt this Thursday, and wanted to share the story here first.
I built this because I journal a lot but hate typing so ReflectAI lets you just talk. Tap, speak for 30 seconds about your day, and it transcribes it, gives you a personalized AI insight, and tracks your mood over time. Every Sunday you get a weekly AI report showing patterns you might've missed.
Built solo with React Native, Firebase, and GPT-4o-mini. It's live on Android already, but Thursday's launch is about getting it in front of more people.
Would love feedback from this community what would make a voice-journaling app actually stick for you long-term? Curious what features matter most.
Congratulations on the launch :) Qq Since it's personal, I'm curious where the AI runs: is transcription and insight generation on-device, or sent to a cloud model? And if cloud, what happens to the data after? Thank you!
@alieksia Hi Anastasiia, thank you so much, and great question!
Right now, transcription and insight generation happen via cloud — I use Whisper for voice-to-text and GPT-4o-mini for generating the reflective insights, since on-device models can't yet match that quality for nuanced emotional analysis.
On data handling: audio recordings are sent securely for transcription and are deleted immediately after processing. The resulting text (your journal entries) is stored securely in Firebase, encrypted in transit and at rest, and is never used to train any AI models or shared with third parties.
I know journaling is deeply personal, so I take this seriously — happy to add more granular controls (like local-only mode or entry auto-delete) if that's something the community wants. Really appreciate you asking! 🙏