Helicase: Multi-Agent LLM System for Supply Chain Knowledge Graphs
A research paper introduces Helicase, an autonomous multi-agent LLM system designed to construct uncertainty-guided knowledge graphs for supply chain analysis. Unlike simple one-shot queries, Helicase addresses structural inference problems requiring multi-hop reasoning across fragmented web resources, such as tracing Tesla components to Australian lithium mines. The system is uncertainty-aware, providing calibrated confidence in its outputs based on source quality and reasoning consistency. The paper is available on arXiv under ID 2605.26835.
Key facts
- Helicase is an LLM-based multi-agent system for supply chain knowledge graph construction.
- It handles structural inference problems requiring multi-hop reasoning.
- Example query: 'Which Tesla components use lithium from Australian mines?'
- The system is uncertainty-aware, providing calibrated confidence.
- Confidence is traceable to source quality and reasoning consistency.
- The paper is published on arXiv with ID 2605.26835.
- It addresses the gap in autonomous knowledge synthesis from fragmented sources.
- Helicase uses autonomous construction and analysis of dynamic knowledge graphs.
Entities
Institutions
- arXiv