Plan before the cargo receiving window collapses.

The cargo receiving window is the period during which a terminal will accept your loaded export container against a specific vessel. It is bounded by the Earliest Receiving Date (ERD) on the front end and the CY Cut on the back end. If your container arrives before the ERD, the terminal turns it away. If it arrives after the CY Cut, it does not make the vessel.

Three reasons. First, when the vessel ahead of yours runs late, the terminal compresses your window to recover yard space. About 70 percent of what you see in the carrier feed is operational noise of this kind, not commercial information. Second, the destination terminal does not know the actual ETA until the vessel sails from the previous port, so it publishes best-guess dates until then. Third, carrier and terminal portals pull from different systems updated at different times, so they often disagree.

VSR shows what changed. Ava tells you when the plan no longer holds. VSR is the system of record. Ava is the system of decision. VSR alone removes the manual checking pattern of carrier portals, vessel schedule PDFs, and spreadsheets. Ava goes further and removes the static-buffer pattern, the "two extra days just in case" rule that fails because the buffer you actually need changes weekly. Most operators need both, in that order.

Production cycles cannot be compressed to match a moving window. Fumigation is days, not hours. Lab certification has its own clock. Equipment availability depends on the drayage provider's chassis pool. Ava's dropoff recommendation factors all three: it asks not just whether the window is stable, but whether your fumigation start time, your lab cycle, and your equipment availability for this booking still let you commit safely. When the answer is no, the alternative is calibrated to the constraint that bound.

No. TMS handles the contract, the booking, the documents, the invoice. VSR and Ava handle the decision-timing layer that sits on top of those: when to act, when to hold, when to reroute. They feed each other. The TMS tells Ava which bookings are open. Ava tells the TMS user which ones need attention this morning.

For most exporters, the first useful report (a daily VSR for the terminals you select) lands within a week. The Ava decision layer typically reaches confidence in 2 to 4 weeks once we have enough booking history to calibrate the per-customer behavior model. There is no rip-and-replace. Most teams run Ava in parallel for the first month, watch where the recommendations would have caught a roll or a missed window, and decide for themselves.

Two ways to answer this honestly. Detection accuracy (did we catch the change in the source data) is high: if a change is published by the carrier or the terminal, we surface it within minutes, and we now cover every US terminal. Outcome accuracy (did the recommendation prevent the bad outcome) is what matters, and Safe-to-Commit lands correctly about nine times out of ten - roughly 92 percent. Where a specific lane is still building history, the system stays more conservative and surfaces the uncertainty rather than hiding it. Every recommendation comes with a record of what you knew when you decided, so if it turns out wrong, you have the trail to learn from.