JARVIS Nexus

Project overview:

JARVIS-NEXUS is an experimental, offline first autonomous AI assistant.

It acts as the central hub connecting various localized sensory modules including Speech-to-Text (STT) for listening, Text-to-Speech (TTS) for vocal responses, and a custom long term memory management system.

The goal is to create a fully private, voice interactive AI that runs completely on local hardware without relying on cloud APIs.

Project status (Early Alpha / Legacy):

Please note that this repository contains the very first proof of concept version.

The codebase is highly experimental, has known architectural limitations, and contains a few bugs.

It is currently serving as a baseline.

I am actively developing a completely rewritten, highly optimized, and stable version from the ground up, which will be released soon.

Tech Stack:

Python, Local LLMs, Whisper (STT), Piper (TTS).