AI Framework Catches Supply Chain Risks 3 Days Earlier

Enterprise AI researcher Nikhil Reddy Donapati has developed a framework designed to detect supply chain risks as they emerge, rather than after disruption occurs. The approach combines Failure Mode and Effects Analysis (FMEA) with large language models to provide both structured risk scoring and situational context that helps operations teams identify the earliest warning signs. Testing on more than 180,000 supply chain records demonstrated strong recall for catching developing disruptions and identified specific inventory and delay thresholds that act as concrete tripwires for monitoring.

The framework's significance lies in timing. Most existing tools report on events that have already happened, leaving limited options for low-cost remediation. By catching problems three days earlier, supply chain teams can choose faster, cheaper responses instead of expensive emergency fixes.

The six-layer architecture moves from raw operational data through risk scoring and reasoning to actionable guidance that operations professionals can immediately use. This advancement addresses a critical gap in supply chain visibility: the need for meaningful warnings with enough lead time to enable proactive response. While the system is not designed to catch every issue, its focus on surfacing significant early signals with clear explanations gives teams the time advantage they need to choose among better alternatives.