Bio Controler

A greenhouse is a map, a calendar and a lab at the same time. When those three live in different places, the team can’t tell if a treatment actually worked.

• An inspection note isn’t tied to a specific spot, so it’s hard to find it again.
• This week looks fine on its own; the trend only shows up when you compare weeks.
• A treatment is logged as a separate form, cut off from the problem that caused it.
• Pest names and doses differ from person to person.
• Hives are hardware with their own failures, and nobody sees them in time.

• Took the domain apart: how a complex is built, how scouting goes, how treatments get logged.
• Checked the product logic with domain experts and the dev team at every step.
• Followed one problem end to end: found in a cell, treated, checked a week later.

• What matters is whether the whole loop runs: the greenhouse is set up, scouting happens on time, a found problem turns into a logged treatment, and someone checks the result.
• For B-Control we also look at whether schedules run, how many hives are online, and how fast the team reacts to odd activity.

• The map, the weekly comparison, treatments, references and hives work as one product, not five modules.
• The hard part wasn’t pretty screens. It was flows a busy agronomist understands at a glance.
• I carried those decisions all the way through development with the team.