Sustainable Systems Lab builds AI-enabled platforms and evaluation frameworks that bring clarity to complex, first-mile systems: supply chains, ESG datasets, and the AI models increasingly used to analyze them.
Supply chains, sustainability data, and the AI systems built on top of them all break down in the same place: the first mile. Smallholder networks, origin communities, and fragmented field data don't fit neatly into the knowledge graphs and dashboards built to represent them. The result is data gaps that create real compliance and decision risk — and AI models that fail quietly, producing hallucinations and misclassifications that go uncaught until they've already shaped a decision.
We structure fragmented data across producers, logistics, and buyers — connecting primary producers to global markets and giving previously siloed actors visibility into their own network. Our work has supported coordination and market access across 107 seafood supply chains.
We identify where AI models break down — hallucinations, misrepresentations, misclassifications — and design structured frameworks to catch and correct them before outputs reach decisions. Applied across NLP pipelines, ESG datasets, supply-chain analytics, and spatial systems.
We link supply chain actors to land cover change, ocean systems, and environmental data — from local fishing communities to global commodity flows — surfacing the gap between what's reported and what's actually happening on the ground.
We use cultural consensus methods to measure whether actors genuinely share standards and practices, or only appear to. Applied to certification, ESG alignment, and supply-chain governance, this work predicts where enforcement gaps will emerge before they do.
Our approach combines three capabilities that trade flow data alone can't provide: spatial intelligence, consensus analysis, and human-in-the-loop AI evaluation. We know what first-mile data actually looks like versus what gets reported — and we build systems that account for that gap rather than papering over it.
Systems scientist and applied AI researcher with a background spanning interdisciplinary ecology (PhD, University of Florida), ESG analytics leadership at NYU Stern, and field research across Chile, Brazil, Bolivia, Indonesia, Madagascar, and Sweden. Fulbright Scholar, NASA Fellow.
Full bio & portfolio →Let's talk about where your first-mile data is breaking down — and what a structured evaluation could catch before it costs you.