Unlock Knowledge with
Local AI
Ingest your private PDFs and documents. Vectorize locally. Chat without limits using LangChain, Qdrant, and Ollama. 100% Free. No API keys required.
Queue A PDF For Ingestion
Upload a PDF and Laravel will queue a background job that copies it into
scripts/data/uploads and runs python3 main.py ingest.
Current pipeline support is PDF only. If you need .docx ingestion, extend the Python loader
first.
Features
Everything you need for Local AI
A complete toolchain orchestrated gracefully. Built beautifully with Laravel, orchestrated heavily via Python LangChain workflows.
Local Vectorization
Process thousands of PDFs locally using HuggingFace embeddings running directly on your CPU. No cloud dependencies required.
Qdrant DB Filtered Search
Automatically category tagging based on folder structure. Leverage the high-performance Qdrant DB to find exact contexts rapidly.
Ollama LLM Integration
Securely formulate intelligent answers by passing retrieved context directly to local models like Llama3 and Phi3 via Ollama.