Sociological studies typically analyze how the use of data produces the social order. Often overlooked is how social (and ecological) systems are ordered so as to be “seen” and governed with data. Drawing on 24 months of fieldwork in Kenya’s “Silicon Savannah,” multi-sited ethnographic case studies, 112 interviews, and full-cycle research methodology, this study documents digital transformations in fresh produce value chains to analyze how platforms remake social-ecological systems in the service of data-driven governance. I make two specific contributions. First, I show how the social organization of markets characterized by extensive relational work limits capacities for datafication. Second, I show how platforms overcome these limits by first constraining possibilities for interaction and then recruiting, shaping, and selecting market actors optimized to these constraints. In the case analyzed, we see how this unfolding, iterative process reconstitutes both the network structure and network composition of actors along the produce value chain, often displacing small farmers and traders with large-scale corporate actors. Implications for inequality, inclusion, and human-environment interaction in data-driven markets are discussed.