LLMs hallucinate regulations and invent transport rules. Lyverto's API grounds your agents in a semantic logistics knowledge graph — query it, subscribe to events, and get answers with the reasoning attached.
// Ask the graph, get grounded reasoning
const answer = await lyverto.reason({
query: "Can WH-A store UN3480 lithium cells?",
context: { customer: "acme" }
});
answer.verdict // "no"
answer.because // [
// "UN3480 → Class 9 dangerous goods",
// "WH-A lacks DG storage certification",
// "WH-C qualifies, 12km away" ]
answer.sources // graph nodes + policies cited// SURFACES
Traverse entities and relationships — commodities, lanes, regulations, capacity — with typed results, not string soup.
Ask operational questions in natural language; get a verdict, the reasoning chain and every source cited. No hallucinated regulations.
Subscribe to shipment.created, customs.hold, temperature.alert and more. Build agents that watch and act — webhooks or streaming.
Claude, GPT or your own stack — connect via Model Context Protocol and your assistant gets logistics expertise as tools.
// WHY BUILD ON IT
Modeling logistics — dangerous goods, customs regimes, equipment compatibility, multi-modal routing — takes domain experts years. It's already in the graph, maintained and versioned. You build the product on top.
Bring a concrete API, graph or MCP use case and define the integration boundaries with our product team.
Discuss API partnership