Spherecast

Spherecast launched Agnes with a deliberately narrow MVP focused on a single supply chain decision: production allocation across existing facilities for one product category. ​​‌‌‌‌‌‌‌​‌‌​​‌​​​​​​‌‌​‌‌‌​​​‌‌Rather than building the full simulation engine across factories, warehouses, and channels simultaneously, they started by solving the most painful bottleneck their CPG customers faced—deciding which plants should produce what. They shipped the initial version in eight weeks, deliberately excluding demand forecasting, logistics optimization, and multi-product scenario planning. This constraint forced them to integrate deeply with one customer's ERP system and prove the core value proposition: could Agnes actually recommend better production decisions than their planners? Early validation came quickly when their pilot customer reduced production changeovers by 23% in the first month, validating that the underlying optimization logic worked. However, this narrow scope also meant they couldn't demonstrate the full "programmable goods flow" vision, limiting their ability to close larger deals initially. The execution approach proved wise—it generated proof points fast enough to secure Series A funding before competitors entered the space, though it delayed their expansion into adjacent supply chain decisions by six months.