Our growers are not going to log into anything. Is that a problem?
Some will and some never will, and the build should assume both. The office keeps working exactly as it does now — tickets in, settlements out — and the online view is an extra for the growers who want it at ten at night. If in two years nobody uses it, you have lost a small part of the build, not the whole thing. We would rather say that up front than pretend adoption is guaranteed.
We are a one-chair shop. What would you actually build?
Almost certainly the cancellation problem and nothing else. An empty chair on a Tuesday is money you cannot recover, and the fix is a waiting list that already knows who can take that exact slot, so the offer goes out the minute the gap appears. That is a small fixed-price piece of work, and if a booking product on the market does it for forty dollars a month we will tell you to buy that instead.
The district consolidated in 2015 and we still have two of everything. Can software fix that?
It cannot fix the politics, but it can stop the duplication being manual. Where consolidation hurts administratively is that the same fact — a pupil's transport, a team roster, an eligibility check — gets maintained twice and drifts. Holding it once and reading it in both places is straightforward work, and it is usually where the first hours come back.
Who is allowed to see records about our children?
Only named staff, and every look is recorded. We build that first on anything involving a school, before any of the convenient features, because a district cannot answer a parent's question about who saw what unless the system was built to answer it. It also means access ends when someone leaves, rather than lingering in an old account.
Where does AI stop and ordinary programming start in a build like this?
Reading a scale ticket, indexing a donated photograph, and matching a form to the right household when the surnames differ — those are model jobs. Applying a contract, printing a settlement, sending a reminder and filling a slot are plain code and always have been. We separate them in the quote so you can see exactly what you are paying an AI premium for, which is usually less than people expect.
The museum has no budget at all. Is there a version of this for us?
Yes, and it is deliberately small. One catalogue, one events page, one volunteer sign-up, handed over so the historical society in the Woodin house owns the account and can run it without us. We quote a fixed figure for that and there is no retainer. A village collection that is searchable is worth more to the town than a bigger system nobody maintains.
Is our information kept in this country, and can we get it out?
Yes to both. The team is US-based, we work under an NDA when you ask, and everything you put in comes back out in ordinary formats whenever you want it. Grower positions and pupil records are not the kind of data anybody should hold hostage, and we put the export commitment in the contract rather than in a sales conversation.
How do we tell whether it was worth the money six months later?
We fix the measurement before the build starts, because measuring afterwards is how projects get to claim anything. For an elevator that is settlement corrections per season. For a shop it is empty chair hours. For a district it is forms still outstanding on the deadline. One number, gathered the same way before and after, and you can check it without us.