Real cases

Industrial artificial intelligence in 100% agave tequila distilleries

A 100% agave tequila distillery is not a volume factory: it is a process factory. What decides the outcome is not line speed but a handful of fine judgments — the autoclave cooking curve, how the tank started, where heads and tails were cut, how long the barrel had — made by people with years of craft behind them. And almost every one of those judgments is recorded today on a sheet of paper, in a notebook, or in a spreadsheet in a shared folder.

See the 12 cases, one by one ↓

That creates two problems which, in this sub-sector, always show up together. The first is economic: agave is 60-70% of variable cost, and yield — liters at 40% per metric ton — is known in aggregate at month end, once the batch is already in barrel and there is no way left to tell why it came in low. The second is traceability: appellation certification is batch by batch, and its evidence lives scattered across paper, spreadsheets, the ERP and the memory of three people. When a plant also bottles for other brands, that same disorder turns into contractual risk.

This page holds the 12 canonical IRIS cases applied to this sub-sector: the five Connect capture modes, the two Connect-to-ERP integrations with direct license impact, the coder integration, the three Edge vision cases on the line, and the agentic case that closes the loop by reconciling every kilo of agave with every bottle. All of them start from the same principle: humans in command. The AI captures, organizes and flags; the decisions stay with the master distiller, the quality manager and the management team.