Derived Insights

Fundamental Intelligence is Kpler's foundational data layer – integrating, connecting, and analysing data to provide clean, asset-level insights across global physical trade. It covers cargo flows, vessel tracking, oil inventories, refinery activity, power and gas, agriculture, metals, dry bulk, and containers, spanning 900,000+ physical assets tracked in real time.

Kpler combines the world's largest terrestrial and satellite AIS network with satellite imagery, drone data, port intelligence, and a global network of 245,000+ human contributors. Over 1.3 billion AIS signals are processed daily, alongside 10 million+ commodity signals, giving users a continuously updated view of where cargo is, where it's heading, and why.

Kpler tracks traded assets across crude oil, clean and dirty petroleum products, LNG, natural gas, LPG, chemicals, agriculture, metals, minerals, dry bulk, and containers. On the infrastructure side, it monitors 1,000+ refineries, 330,000+ km of pipelines, 9,800+ ports, 19,300+ terminals, and 12,000+ oil storage tanks globally.

Kpler's data is continuously validated by a global team of 70+ market analysts and back-tested against customs data and other independent sources. Key accuracy benchmarks include 99.9% LNG accuracy vs customs data, 98.5% accuracy for crude exports from the world's top five exporting countries, and 99% correct geospatial shapes and hierarchical relationships across ports, berths, and anchorages.

Kpler draws from over 255,000 proprietary sources and 105+ private sources, combining satellite imagery, terrestrial and satellite AIS, drone data, customs records, port data, weather, outages, news, and community intelligence. Raw data is processed using proprietary AI and machine learning, with expert analyst validation applied throughout to ensure accuracy and reliability.

Kpler has 15+ years of longitudinal trade and vessel movement data. This historical depth is critical for back-testing quantitative models, identifying structural trade patterns, and providing a foundation for AI and machine learning that cannot be recreated by newer entrants to the market.