About Kumi Analytics

Kumi Analytics' story goes back to 1992 when its Founder and Managing Director, Clinton Libbey, joined a small startup in Pasadena, California—Core Software Technology (CST). CST's heritage came from NASA's Jet Propulsion Laboratory and Caltech. While at CST, the CTO, Dr. Robert Nathan, who is recognized as one of the inventors of digital image processing, often stated that satellite imagery is one of the richest sources of information but it is very difficult to extract as unsupervised classification was not reliable—or at that time even possible given the complexity of the data source. While working his way up to Vice President, Clinton was involved in launching the first global satellite imagery browsing system, and helped build a global ground station network in support of the first international high resolution satellite imagery joint venture (of which CST was one of the founders).

He was also part of the Environmental Intelligence Application Program (EIAP) at The MITRE Corporation—the first use of very high resolution imagery for the long term analysis of environmental sites of interest.

Twenty years later, while working on Global Strategy for Maps and Earth at Google, Clinton began to see unsupervised classification—now referred to as machine learning or deep learning—with satellite imagery become not only viable but, when coupled with cloud technologies, scalable. The technology finally caught up with the vision of one of the inventors of that same technology, and the possibilities are now endless.

Kumi Analytics was initially focused on leveraging satellite imagery to provide financial institutions insights on the economy and developed environmental applications in order to give back to the world and protect the planet for future generations. After building a proof of concept for the Singapore Government's Emerging Stronger Together Taskforce on green finance, the company shifted its primary focus to supporting the environmental market and integrated the passion of protecting the environment with the science of remote sensing.

Our deepest technical edge came from mangroves—the structurally complex, tidally inundated, perpetually cloud-covered ecosystems that are among the hardest on Earth to measure accurately. Delivering audit-grade dMRV in mangroves proved our methods under the toughest conditions; that same precision now supports forestry, REDD+, soil carbon, grassland monitoring, and geospatial compliance across nature-based solutions portfolios.