Skip to main content

Privacy Overview

RAG-DocBot is built with privacy as a core design principle, not an afterthought.

On-Premise by Design​

RAG-DocBot runs entirely on your own infrastructure. There are:

  • No external API calls — all AI inference is handled locally by the inference service using your GGUF model file.
  • No telemetry — no usage data, error reports, or analytics are sent anywhere.
  • No SaaS dependencies — the system operates independently of any cloud service.

All document data, conversation history, and user information stays within your environment.

Air-Gapped Deployments​

Once the Docker images are pulled and the system is running, RAG-DocBot does not require internet access. It is suitable for deployment in air-gapped or highly restricted network environments.

Privacy by Default​

Out of the box, RAG-DocBot applies privacy-preserving defaults to its logging and data retention behaviour. See Data Retention for details.