@PABLOF7z would you be interested in an associative reasoning engine that works on the edge/browser, in any language, scales to massive datasets and essentially acts as a cybernetic mediator between your client and the NDK? Spent the last year building it as an alternative to LLMs and am prepared to open source, but I think a proper rollout with Nostr implemented would be more strategic than an anonymous git commit. I envision a decentralized internet where you attract information instead of chasing it.