Customer Flow AI
Multi-camera vehicle tracking under a single ID across the entire drive-thru journey — actionable above-store insight at a fraction of competitors' cost.
Many customers wanted computer vision at their stores — both in-store and in the drive-thru — to understand throughput and service times.
After vetting and selecting TraceVision as a partner (based on their proven work in the sports space), I co-developed a proof-of-concept for Dutch Bros and other drive-thru customers that tracks vehicles across multiple cameras under a single ID throughout their entire journey. This enabled vehicle counts and speed-of-service times — broken down by stage (ordering, window, line stack, etc.) — plus detection of drive-offs and lane abandonment, all surfaced in dashboards. I personally built a session-review tool so customers can filter by any parameter (e.g., longest service times) and watch the associated video, making the data actionable rather than just another set of numbers.
Delivered actionable above-store insights at a fraction of competitors' cost.
Cloud-based processing with footage passed through a Raspberry Pi at the edge — substantially cheaper than competitor solutions.