Case Study
COSA: field data from coffee farms, ready for analysis
Sustainability Standards · Data Engineering · Staff Augmentation
The Committee On Sustainability Assessment (COSA) is a global consortium that measures the sustainability of agricultural supply chains, working with partner organizations around the world. Its work runs on data those partners entrust to it.
In coffee, much of that data is collected in person. Field teams visit producers on their farms across South and Central America and record what they find there. For two years, a Cloudacio data engineer worked as part of COSA’s team on the step between those visits and the analysis the organization delivers to its partners.
Many of the producers COSA works with are small farms that keep few formal records. A field researcher collects what the producer has, in whatever form it exists: a spreadsheet from a cooperative, answers entered in a survey, figures written on paper. Two visits to similar farms can come back with different columns, different units and different names for the same thing.
Before anyone at COSA can compare one region with another, that material has to be cleaned, matched and put into a single structure. New data arrives with every field campaign, so the work never finishes, and it takes someone who knows both the engineering and the habits of each source. COSA wanted that capacity inside its own team for as long as the field work continued.
Cloudacio placed a data engineer with COSA as a member of its team, working in its tools and on its priorities. The engineer built and maintained the pipelines that take field data as it arrives, validate it, convert it to common units and definitions, and load it into the datasets COSA’s analysts use.
The length of the engagement mattered. Every source has its own way of reporting, and the rules for handling each one accumulate over time. An engineer who has been through several campaigns knows which gaps are errors and which are simply how a given cooperative reports, and writes that knowledge into the pipeline so the next campaign needs less manual repair.
COSA’s analysts receive field data in one consistent structure, whatever shape it had when it left the farm. Normalization happens in the pipeline, so analysis starts from data that is ready to use.
COSA also worked with Cloudacio on its SOC 2 Type I and Type II.
"The data our field teams bring back from coffee farms never looks the same twice. Cloudacio’s engineer joined our team, learned our sources, and turned that into data our analysts can work with."Pavel Muñoz, Director of IT at COSA