SBIR-STTR Award

CHAIN is technology for transforming how supply chains see the future. CHAIN machine learns to anticipate change on the ground – shifts in consumption and utilization – and recommends exactly the right amount of supply for each physical location. CHAIN learns from routinely collected data, atmospheric data, satellite imagery to understand the physical infrastructure that increases or decreases access, and leverages human-in-the-loop machine learning - technology for learning directly from what expert users share in an unstructured format [text messages, images] – to gain access to rare, real-time insight, and gather points-of-view that wouldn’t otherwise be captured. CHAIN analyzes hundreds of dimensions of data to find elements and data-structures that have predictive power – tests these models in a blinded fashion where the outcomes are removed so that the technology doesn’t know the right answer – and scans patterns in data in real-time, looking for layers of clues that suggest how and when change will occur. CHAIN optimizes for supply chain performance by building on predictions, incorporates the mathematical constraints learned from the supply chain [i.e. cannot deliver commodity x to location y in less than 24 hours] and recommends how and what supplies shoul