Case Study: Sylvera achieves rapid, large-scale mangrove annotation to verify carbon offsets with CloudFactory

Sylvera brings transparency to the carbon-offset market with AI

Sylvera, a London-based climate-tech startup that rates carbon-offset projects, needed high-resolution, time-series labeling of land use—especially mangroves—to verify that projects were delivering promised carbon sinks. Rather than building an in-house team or relying on coarse public datasets, Sylvera engaged CloudFactory’s Computer Vision Managed Workforce for a customized annotation solution to produce the millions of labeled satellite images required.

CloudFactory provided a dedicated, trained annotation team (integrating Azavea’s GroundWork where needed), delivering 15,000+ square miles of labeled mangrove imagery—an area nearly 25 times the size of London—in just two months and saving Sylvera more than 500 hours of work. The high-quality labels enabled Sylvera to train deep learning models that improve monitoring accuracy (in some cases correcting ground-truth approximations), accelerating project verification and scaling their transparency efforts.

Sylvera

Virginie Bonnefond

Machine Learning Engineer