A village of eleven hundred people paying for custom software — can that possibly be worth it?
Sometimes it plainly is not, and we will say so on the first call. The test we use is whether the thing you want fixed costs somebody more than a few hours a week or has already cost you money once. A festival committee losing a Saturday to chasing vendor certificates passes that test. Wanting a nicer-looking website usually does not, and there are cheaper ways to get that. Worth saying, though: between the Railroad Avenue elevator, the Lewis Street inputs trade and the pipeline work in Pleasant View Township, this village runs more regulated operations per head than a lot of towns five times the size.
Our elevator already has grain accounting. Would you replace it?
No, and we would push back if you asked us to. Grain accounting packages handle contracts and settlements properly and rewriting one is a bad use of money. Where we help is at the edges — capturing tickets faster during the busy weeks, or giving growers a view of their own position so the office phone rings less. That connects to what you own; it does not replace it.
Can software decide what to recommend for a field?
It should not, and we will not build it that way. Rate decisions, product choices and substitutions belong to a licensed agronomist, full stop. What software does well is check that what was ordered matches the acres and the label, and flag it when it does not. That is a safety net under a professional, not a replacement for one.
The festival committee changes people every couple of years. Will this survive that?
Only if it is built to. So the accounts are in the committee's name, not an individual's; the records export in ordinary formats; and we write the handover notes for someone who has never seen the system before. A tool that only works while one particular volunteer is still on the committee is worse than a paper folder, because at least a folder can be picked up.
The village board would have to approve any spending. What do we show them?
A fixed scope, a fixed price, and the number we expect to change — billing corrections, days to answer a resident, whatever fits the job. Public money should come with a check you can run afterwards, and we would rather write that check into the brief than argue about value at the end.
Who has the records — us, or you?
You do, entirely. Grower files, resident accounts, vendor lists and event history sit in infrastructure registered to your business or your village, and a complete export is available whenever you ask. Nothing you put in gets used to train anything, and none of it is shared. That goes in writing before you spend a dollar.
How much of what you would build is genuinely AI?
Less than half, usually. Reading tickets, certificates and permits and pulling the numbers off them is real machine work and saves hours in a busy season. Billing, contracts, bookings and rosters are ordinary software and we build them that way, because they need to be exactly right every time and there is no reason to pay a premium for arithmetic.
What is the right time of year to do this?
Winter, for almost everything here. A build takes four to eight weeks, and going live in the middle of harvest or the middle of a spring rush is a mistake. If we are doing something for the festival, we want it running by spring so the committee has a full cycle to get used to it before the park fills up.