CCU · Beverages
From 4,000 to 45,000 samples a month, each one traceable
CCU needs to know how long a returnable bottle takes to come back to the plant in order to size its bottle fleet. Samplers used to write down label codes by hand and send them over WhatsApp. Now they take a photo and computer vision does the rest.
AI capture
11×
more samples with the same team
Results
11×
more samples with the same team
45,000
samples a month, up from 4,000
4
plants in Argentina
The problem
A mobile app that captures the label codes, loads them into a database and calculates the turnover of each sampled bottle as an indicator, to estimate the bottle fleet that the sales volume requires.
What Tracy does
A capture app with computer vision, image approval and export to the team's dashboard.
- A photo of the label from a phone, instead of writing down every code.
- Automatic reading of bottling codes with computer vision.
- Every sample traceable: who, where and when, straight to the team's dashboard.
Status today
Fully operational in 2026 at all four plants in Argentina.
