Each of the 5 properties decoded with real examples from AfriPay, MediLink, and Lagos Fashion. The framework that scores any AI deployment in Africa — and shows you exactly which letter is broken.
One letter at a time. Each property paired with a real African deployment that nailed it — or failed it.
The agent acts independently within a defined scope. No human in the loop for routine queries. Escalates only what humans need to handle.
Anchored to your actual business data — products, prices, hours, policies. Cannot hallucinate beyond what's in the knowledge base. Sources every claim.
Works on 2G. Works on a $40 Android. Works when the network drops mid-message. Degrades gracefully — never fails silently.
Speaks the customer's actual language — Kirundi, Swahili, Wolof, Pidgin — with cultural fluency, not Google-Translate output. Native means trained, not translated.
The customer understands the decision — in their language, in plain words. Not stack traces, not engineering logs. Real transparency = the user knows why.
A·G·E·N·T is not a checklist of nice-to-haves. It is a system. An autonomous bot that's not grounded hallucinates. A grounded bot that's not edge-ready dies on 2G. A native-speaking bot that's not transparent loses trust on the first ambiguous answer. Score on all five — or don't ship.
Lesson 2.3 is the hands-on application: a 5-column, 12-cell worksheet that scores your deployment letter by letter and tells you exactly what to do next: ship, fix, or redesign.