Aiden's current development-board architecture has no Aiden-hosted backend. Screenshots, voice, and prompts go to the model, speech-to-text, and text-to-speech endpoints the user configures, not to a backend Aiden operates. The firmware is open source, so the architecture is fully inspectable, and self-hosting or local-model configurations are supported.

That said, Aiden also supports persistent context, preferences, and skill optimization, so a deployment can retain certain information locally. The article lays out exactly what's confirmed today (no hosted backend, bring-your-own-provider architecture) versus what hasn't been publicly documented yet (retention periods, deletion workflows, logging defaults, encryption specifics) — without overstating either direction.

Full article: aidenai.io/blog/does-aiden-store-your-dataFirmware: github.com/AidenAI-IO/aiden-firmware