The Logbook

A self-sustaining system is not declared — it is built subsystem by subsystem, with the data that each stage produces. The nodes of this series document that process in real time: what works, what fails, and what is still a hypothesis.

Operating with a single AI model is convenient until that model fails, censors a task, or charges you three times for reasoning that another does just as well. This resource documents the orchestration architecture I use in production from the Ambalá-Calambeo corridor: eleven models with fixed roles, a failover matrix, and prompting rules that prevent the AI from inventing business rules.

There is a difference between using AI and working with AI. The former is loyalty by default. The latter is architecture by choice.

If each subsystem of the farm —meliponary, solar, water, agroecosystem— is an instrument of an orchestra, Home Assistant with InfluxDB and Grafana is the score that allows us to see, for the first time, how they all sound together — not just listen to each instrument separately.

This node explains how to properly size an off-grid system for a farm with automation, what Law 1715 says exactly about each of its four benefits, where the real risks of misapplication lie, and what it means.

*A tutorial ends when the flow works the first time. Production starts exactly there.*

Most AI stacks built in 2024 and 2025 were fragile. They didn't know it because the pressure had not yet arrived.