This is not a science project; the product works and demand is real. The main risks are that
dilution arrives before cash generation, acquisitions add complexity faster than margin, and larger platforms turn conversational AI into cheap bundled infrastructure. The critical proof points are gross-margin repair, lower burn, successful
LivePerson execution or clean fallback, and evidence that customers are buying workflow outcomes rather than commodity voice minutes.