In 2013, we assumed counting people was a solved problem. We looked for technology that could do this and found hundreds of products — stereoscopic, ultrasound, thermal, PIR, AIR, low-res cameras, wifi tracking, bluetooth, etc. We bought them all and experimented. Everything turned out to be cheap and inaccurate or expensive and invasive. Even the pricey stuff was terrible at people count. What began as a weekend project has become the better part of a decade working to build the standard platform for counting people — accurately, anonymously, and in real-time.