OriginTrail Decentralized Knowledge Graph (DKG)
Workflow Diagram
Task
OriginTrail DKG enables people and agents to turn data into verifiable structured interconnected knowledge that remains traceable to its source.
Intent
The AI system combines symbolic AI (a DKG) with GenAI to create a resilient decentralized AI infrastructure with source traceability.
AI Workflow
Input
A user submits data (e.g. PDFs) to the AI system.
Process
Agents transform data into a knowledge graph draft (nodes and connections) for the DKG; roles may include coordinator or ontology expert.
Output
A review gateway (a human or an agent) can approve/adjust the DKG before publishing; once published the DKG gains cryptographic source verifiability and links with other knowledge on the network.
Human Oversight Level
Conditionally Autonomous AI
Institutional Oversight Examples
- •User/community-led verification of the knowledge graph (ad-hoc practice)
- •Open-source code (organization policy)
Risk
It is possible for agents to make a mistake while building the graph; an approval agent or human may miss an error and become over-reliant.