Search Results

Overview This chapter documents the AI inference, voice processing, and workflow automation services running on Centerpoint. All GPU-accelerated workloads leverage an NVIDIA GeForce RTX 5080 connected to the Mini PC via OcuLink, providing 16 GB of GDDR7 VRAM o...

Overview Ollama is the local large language model (LLM) inference backend for the homelab. It serves models via an OpenAI-compatible REST API and is consumed by Open Web UI, PaperlessAI, and any other service that needs LLM inference without sending data to ex...

Overview Open Web UI is the primary chat and AI management interface for the homelab. It provides a ChatGPT-style web frontend connected to the local Ollama backend, with support for conversation history, model selection, RAG (document chat), image generation,...

Overview Faster-Whisper is the speech-to-text transcription service for the homelab. It runs OpenAI's Whisper model via the faster-whisper library (CTranslate2 backend), which offers significantly faster inference than the original Whisper implementation at e...

Overview Kokoro is a high-quality, locally-hosted text-to-speech (TTS) service. It exposes an OpenAI-compatible TTS API, making it a drop-in replacement for external TTS services in any application that supports the /v1/audio/speech endpoint. It runs the Koko...

Overview Riffado (formerly known as OpenPlaud) is a self-hosted AI audio and podcast management application. It handles audio file storage, playback, and AI-assisted processing within the homelab ecosystem. Audio files are stored on the Ceph OSD volume ( 1TB_V...

Overview PaperlessAI is an AI companion service for Paperless-NGX that automates document classification. It monitors the Paperless document inbox and uses a local Ollama model to automatically assign tags, document types, correspondents, and titles to newly i...

Overview Open Notebook is a self-hosted AI research notebook application — a local alternative to Google NotebookLM. It allows structured AI-assisted research sessions where sources (documents, URLs, text) are ingested into a notebook and queried via an LLM fo...