Imagine a densely packed crowd with a common voice, such as people in a candlelight vigil or a protest. We envision an innovation through mobile computing technologies to empower such crowds. We present Card-stunt as a Service (CaaS), a mobile service allowing them to instantly create a massive collective visualization by simply holding their phones up. The key challenge is to achieve instant, infrastructure-free, decimeter-level localization of individuals in a massively packed crowd, within a low latency. CaaS addresses the challenges by mobile visible-light sensing and scalable constrained optimization. Within less than a minute, each person is given individualized pixels so that they can do their part in the overall visualization.