Descartes Labs

Descartes Labs operates with a scientific leadership team led by CEO and co-founder Mark Johnson. Investment decisions appear to flow from the firm's proprietary machine-learning models, which generate trading signals from satellite data. There is no separate CIO structure publicly disclosed.

Structurally, it is a hybrid. The firm sells its geospatial analytics platform to enterprise customers for supply-chain monitoring and sustainability tracking. But its founding team built the technology to exploit trading opportunities in commodity markets, and Descartes Labs has historically used its own signals to inform principal trading in agricultural commodities and related derivatives.

The firm's satellite-based machine learning models are tuned to forecast production in agriculture, energy, and metals and mining. Its models also extend to environmental monitoring for sustainability-linked finance and insurance. While primarily focused on commodities, the platform can detect changes in any physical asset visible from space, from shipping traffic to deforestation.

The firm's core deep-learning capability was developed at Los Alamos National Laboratory, where the founding team worked on national security applications requiring detection of subtle signals in vast satellite datasets. In 2014, they spun the technology into a commercial entity, adapting it for commodity markets and enterprise customers.

Descartes Labs was selected for the Defense Innovation Unit's first cohort, which fast-tracks commercial technology into the Department of Defense. The partnership validates the firm's ability to map and monitor physical systems globally using AI-driven satellite analysis, though the specific defense applications remain largely undisclosed.