Does a mile of old road really generate business?
It generates visits; whether it generates money depends entirely on whether anything tells those people where to eat and how long to stay. That is a solvable problem and a cheap one. It is not a substitute for the farm economy and we would not sell it as one, but a good visitor page and a shared calendar is one of the lowest-cost things a town like this can do.
We haul grain. What would you actually build for us?
Usually the ticket-to-invoice path first, because that is where the money and the evenings go. Photographs of scale tickets and bills of lading read and matched to loads, driver qualification dates tracked next to the dispatch board, and settlement that reconciles without somebody sitting up with a calculator. It is not glamorous and it is the thing that pays back fastest.
Can this run on a phone with bad signal at an elevator?
It has to, so we build for it. Entries queue on the device and send when there is signal, and nothing important depends on a page loading at the pit. We test this on the actual equipment your drivers carry, not on a desk in an office with good wifi, because the version that only works on wifi does not get used.
The city has almost no staff. Is this more than we can maintain?
It should be less, not more, or we have built the wrong thing. The test we hold ourselves to is whether a clerk can run it without calling us, which means plain wording, few settings and no dependency on one person's account. If a proposal cannot pass that test we redesign it before quoting.
What actually needs a human in the loop?
Anything published in your name, anything with a price or a deadline, and anything about a person's qualification to drive. The gate is called SolaceSentry and it holds the item until somebody releases it. The things that do not need a gate — matching a ticket to a load, sorting an inbox — run on their own and we do not pretend they are dramatic.
Where does our information live and can we take it with us?
Under your accounts, and yes. Load history, customer records, visitor sign-ups — you can export the lot in an ordinary format at any point, and the arrangement is written down before you pay us. We have never seen a good reason to make leaving difficult, and plenty of reasons not to.
When should we start if we want it working next year?
For anything visitor-facing, over the winter, so it is live before travelling season. For a hauling or shop build, between harvest and planting for the same reason — there is no useful training week in October. Four to eight weeks is the normal span once mapping is done.
Are you going to try to sell us AI we do not need?
We try not to, and we put the split in writing so you can check. Reading a stack of scale tickets is AI doing real work. A calendar that shows what is open on Saturday is a database and should be priced like one. If a whole proposal comes back with no AI in it, that has happened and it is a perfectly good outcome.