Software built for the conditions it actually runs in.
Intwasa builds and operates seven products across agriculture, education, commerce and document intelligence — and advises on large-scale agricultural development. All of it designed for intermittent data, low-end handsets and users who cannot afford for the software to be wrong.
- 7products built and operated
- 5divisions, from satellite agronomy to advisory
- 3countries of operation
- 2026incorporated in South Africa
An operating company, not a consultancy that also ships apps
Intwasa builds software and runs it. Every product in the portfolio is operated by the same team that wrote it — the satellite pipelines, the classifiers, the payment flows and the servers they sit on. That is a deliberate structure: software that touches a harvest, an examination or a land purchase cannot be handed over at launch and left.
The work is concentrated where the constraints are hardest. A vegetation index is worthless if the farmer cannot open it on a handset with no signal, so the pest classifier runs on the device rather than the server. A past-paper archive is worthless if the data cost exceeds the tutoring fee, so it caches offline. These are not features. They are the reason the products are used at all.
Alongside the products, Intwasa takes on advisory work in agricultural development — environmental and social impact assessment, land capability, and the programme tooling that keeps a multi-party development on schedule. The two sides feed each other: the agronomy that makes AgroSync correct is the same agronomy a fertility study turns on.
Five divisions under one company
Divisions are trading names, not separate registrations. The consumer brands keep their own identities and carry the endorsement quietly.
-
Intwasa Agri
Satellite agronomy and land assessment.
AgroSync·Geo Dossier
-
Intwasa Learning
Examination preparation, secondary through tertiary.
Scholar·Varsity
-
Intwasa Commerce
Marketplace infrastructure for physical retail.
ZimMall
-
Intwasa Intelligence
Applied language models against documents that carry risk.
Stratagem·Soul-Finder
-
Intwasa Advisory
Environmental, land and programme advisory.
ESIA and land studies·Programme tooling and training
What the team actually does
Machine learning
Image classifiers trained in-house and quantised to run offline on the handset, because the field has no signal. Applied language models where the input is a document and the output has to be defensible.
Geospatial
Satellite index pipelines, parcel geometry, geodesic measurement and cartography, from raw scene to a figure that survives a review meeting.
Agronomy
Crop phenology, variety selection, fertility interpretation and export certification — the domain knowledge that makes the software correct rather than merely working.
Platform engineering
Python and FastAPI services, Android applications, and the unglamorous delivery work — self-updating installers, offline caching, resumable downloads — that decides whether software reaches a rural user at all.
Intwasa
In isiZulu and isiXhosa the root -thwasa means to emerge anew. It describes the new moon appearing and the turn of a season — intwasahlobo, spring. It is the right word for a company whose oldest product watches a crop come up from orbit.
Work with us
For advisory engagements, partnership enquiries, procurement documentation or press, write to hello@intwasa.com.