AI Automation en Jeddah
How we work in Jeddah
The calendar of non-working days is a table, not a constant
Writing those dates into the code turns them into something that expires in January and that only someone who knows how to deploy can touch. In a table anyone on the team maintains them, the history is kept, and the automations read it instead of trusting a stale copy.
Some dates are confirmed at short notice
When next year is still not settled in September, the system has to accept a last-minute change without touching code or waiting for a deploy window. That gets designed beforehand: a screen, a permission and a record of who changed what and when.
A reminder is worked out against the calendar, not out of thin air
«Warn three days before» and «retry later» give different results depending on which dates the system discards. If that rule lives spread across five automations, four get fixed and one is left firing messages on a day when there is nobody on the other end to answer them.
An agent should not convert dates from memory
You can ask a language model for an equivalence between calendars: it will answer confidently and now and then get it wrong, and with a date that does not show when you read it. A function with its own test does the conversion; the model writes around the result.
Answering your questions
What people ask us about AI Automation in Jeddah.
If your question is not here, tell us about the project and we will answer with real context.
Ask us directlyReal work
Projects we have built
Full service · All the detail
AI Automation: stack, process, use cases and FAQs