Dokis enables RAG systems to verify source attribution without requiring additional LLM inference calls

Dokis is a tool for runtime enforcement of provenance in Retrieval-Augmented Generation (RAG) systems

Eliminates the need for extra LLM calls to validate that generated responses are properly grounded in retrieved sources

Improves efficiency and cost-effectiveness of RAG pipelines by reducing computational overhead

Available as open-source project on GitHub for developers building fact-grounded AI applications