LORIE

Local Offline Reasoning Intelligence Engine

LORIE is a personal offline LLM system: a unified reasoning and retrieval engine that runs entirely on local hardware. It uses a curated knowledge corpus, embeddings, domain routing, and tool modules to generate detailed, cross-domain answers without the cloud.

Core Features

• Domain-aware router for history, psychology, sociology, medicine - first aid, coding, technology, math, mechanics, animals and husbandry, land and gardening, DIY, engineering basics, ancient history, anthropology, forensics, and more.
• Retrieval-Augmented Generation (RAG) using local documents.
• Deductive and cross-domain reasoning engine.
• Metadata scoring, weighting, and relevance ranking.
• Verification layer (consistency checks and contradiction filters).
• Persona layer that adapts to personal tone, clarity, and reasoning style.
• Local model loader for GGUF, Ollama, and quantized LLMs.
• PyQt6 desktop interface and FastAPI backend for clean separation of UI and logic.

Goals

• Provide ChatGPT-like reasoning fully offline.
• Remain usable during outages, travel, camping, or emergency conditions.
• Serve as a long-term, expandable personal knowledge system.
• Remain fast and responsive on consumer-grade hardware.
• Support ingestion of books, PDFs, notes, images (OCR), and transcripts.

Status

• Active development.
• Current baseline release: v0.1.0.