Tracking the line where tax changes.
Meet SITUS-1, an agentic AI sales tax boundary investigator examining real, geotagged street imagery for mailbox candidates. Those physical clues can corroborate where a transaction lands.

The investigation, in motion.
SITUS-1 is moving through geotagged street imagery, looking for mailbox candidates and calculating their estimated positions.
One real mailbox. One calculated position.
SITUS-1 retains the most recent positive candidate as a field observation. The still, detection box, camera vector, and measurements below all come from that same provider frame and classifier result.
Why locate a mailbox at all?
Sales tax sourcing often starts with where a product is delivered. SITUS-1 explores whether a visible delivery-point clue can corroborate the destination coordinates already evaluated by Prophit.ai’s tax engine.
Establish the ship-to location.
The deterministic engine begins with the transaction’s delivery address and the sourcing rules applicable to that sale.
Observe the delivery point.
A mailbox candidate can provide a visible clue that a delivery point exists where the address and imagery say it does.
Project the clue off the road.
Camera coordinates, heading, image offset, and estimated distance produce an experimental mailbox coordinate with stated uncertainty.
Compare it with tax boundaries.
The projected point can be compared with governed state, local, and special-district evidence before any tax conclusion is served.
From one investigator to a fleet.
A national visualization of the Boundary Agent fleet, with SITUS-1 highlighted at its current scan location.
Because tax follows place.
Delivery destination is a core sourcing signal in sales tax. The Boundary Agent is an experimental corroboration layer alongside Prophit.ai’s deterministic address, rate, and jurisdiction systems. It does not replace them.
Explore the Prophit.ai platform →