- AgriTech & traceability
FAMS client implementations
Configuration, data import, on-site training and custom features for agribusinesses such as Thandizana, Moz Agri, DREAM Agro and Green Agro — with OLIMAX, MHC, Sociedade de Mutuali and Moz Agri on their own pages.
- Period
- 2024–2026
- Status
- In production
- Role
- Technical implementation lead and trainer
- Partners
- NIRAS, GIZ
Problem
A system only helps if the company uses it. Every client arrives with farmer lists in Excel, its own terminology and a different value chain, and field staff need to learn the app on the ground.
Solution
Implementation missions follow the same routine: structure the data for import, configure areas, groups, communities and forms, enable the needed plugins, train office and field users, and validate the system in real operation. Where a client needed it, Edson also built features in FAMS.
Technical decisions
01
Thandizana: repayments and dividend distribution
Thandizana (Manica) gives farmers input credit and pays a marketing bonus. In weekly sessions with the client the rules were validated together — a farmer is only paid if nobody in their group is in default; the repayment is the quantity taken times the campaign value; the bonus depends on kilos sold. The Repayments and Dividends module was completed and its first part is already in use.
02
Green Agro: duplicate warning when creating farmers
Farmer registration requires a GPS-tagged photo as proof of visit, and purchases record the maize moisture level. The July 2026 mission deactivated around 150 duplicate records; rather than cleaning by hand each time, FAMS now warns about likely duplicates at the moment a farmer is created.
03
Moz Agri: configuring for the real operation
Five days in Manica (26–30 January 2026) of configuration and training, with validation in the office, the field and the factory. Later the product register was extended to cattle and goats (male and female), because the client also buys livestock.
04
DREAM Agro: the full cycle before leaving training
In Niassa, training ended with an end-to-end simulation — farmer registration, training attendance, visit survey and purchase — and GPS plotting of a real production area, synced to the server. By July 2026 the team was running the create–edit–photograph–sync cycle on its own.
Images
No publishable screens: the system needs real data to run or is no longer online.
Verifiable results
- Thandizana farmers in FAMS — up from 2,937 in Feb 2026
- 6 329
- Moz Agri farmers in FAMS — up from 72 in Jan 2026
- 4 780
- DREAM Agro farmers in FAMS — up from 1,370 in Feb 2026
- 1 532
- duplicate records fixed at Green Agro
- ~150
What I learned
Installing is not adopting. On one follow-up mission, of about 25 agents, 11 used the app on their device and only 1 synced regularly: the problem was not technical but routine and supervision. The answer was both product and operations — a dashboard with each agent's last sync, usage targets, planned sync windows and hotspot sharing in the field.
SDGs

SDG 2 · Zero Hungerprimary
Each rollout moves farmers from paper and Excel lists into georeferenced digital records, with trainings, surveys and purchases the company and its funders can follow.

SDG 8 · Decent Work and Economic Growthprimary
Configuration, data migration and on-site training professionalise the companies' operations and make farmer credit and payments more predictable.

SDG 10 · Reduced Inequalitiescomplementary
The missions reach areas with poor coverage — in Niassa, Manica and Zambézia — where offline capture and planned sync windows are the only way the data can exist at all.