Four Kites

FourKites operates a supply chain platform that combines a logistics data network, a predictive intelligence layer and a set of AI agents that execute operational work. The company describes its network as covering more than 500,000 carriers, 3.6 million facilities, 2.7 million lanes, 1,600+ global brands and 234 countries, enriched with external signals including customs, sanctions, weather, maritime, geopolitical and commodity data, and states the network has been accumulating data for 11 years and processes more than 3.2 million tracked events per day.

On top of that data, FourKites markets an intelligence layer ("The Graph", FourSight AI and real-time "Digital Twins") built from 116.2 million historical loads, used for tasks such as identifying carriers losing capacity, attributing facility-driven delays and predicting purchase orders at risk of missing delivery windows. Execution is delivered by six named AI agents in production at Fortune 500 companies: Tracy (carrier operations and shipment monitoring), Sam (supplier operations and document processing), Alan (facility operations and appointment scheduling), Cassie (customer operations and delay notifications), Polly (compliance and documentation, proof of delivery, customs validation, PO lifecycle) and Sophie, an agent-building agent that converts plain-language process descriptions into production workflows. The company reports roughly 2.5 million autonomous actions and 2.5 million decision traces per month.

FourKites packages its capabilities as 152 "outcomes" spanning five domains: Supply Resilience, Customer Fulfillment, Site Intelligence, Global Movement and Network Intelligence. Example outcomes include touchless ocean booking, autonomous carrier follow-up, ML-powered truckload ETA, door-to-door ocean visibility, cold chain integrity, computer vision gate automation, yard trailer inventory, ocean detention-and-demurrage auto-disputes and natural language supply chain analytics.

Business model

FourKites sells enterprise software positioned as "Outcome-as-a-Service for supply chain," explicitly stating it does not sell per-seat licenses but instead charges for replaced work and delivered outcomes rather than configurable features.

Commercial engagement is framed around outcomes rather than user seats, with prospective buyers routed to an "Outcome Advisor" rather than self-service sign-up.

Traction

Reported usage includes more than 3.2 million tracked events per day, about 2.5 million autonomous agent actions per month, 116.2 million loads of historical data and six agents in production at Fortune 500 companies. Customer-stated results include response times cut from 90 minutes to seconds at Church & Dwight and a 98.6% reduction in detention costs at two First Solar sites.

Latest developments

The company has repositioned from supply chain visibility toward "AI-Powered Supply Chain Orchestration," with a stated mission of completing one billion hours of operational work via AI agents over the next decade and a digital workforce of six named production agents.