Show HN: Local AI server with persistent memory, RAG and plugins (github.com)

A new local AI server, Server Nexe, has been announced, featuring persistent memory capabilities and a zero-cloud approach that ensures users maintain full control over their data. This platform, released as a Tauri v2 desktop application for macOS and Linux, allows users to run their own AI model while integrating various inference backends like MLX, Ollama, and llama.cpp. Server Nexe stands out for its zero data leakage policy, ensuring that all interactions, documents, and embeddings remain local, enhancing user privacy and security. For the AI/ML community, this tool is significant as it represents a shift towards decentralized AI usage, where individuals can tailor and manage their AI infrastructure without relying on external servers or cloud-based services. The application employs advanced techniques such as retrieval-augmented generation (RAG) to enable persistent memory, allowing the AI to remember context across sessions. Additionally, a modular architecture facilitates easy plugin integration, enabling users to expand capabilities without altering the core system. With over 6,600 tests and built-in security measures like AES-256 encryption, Server Nexe showcases robust production readiness, despite its initial status as a personal project.