South African Enterprise AI in Practice
The AI features that impress on conference stages usually rest on years of unglamorous work: re-platforming after outages, owned domains, trustworthy data and redesigned processes. This four-part series draws out those patterns using fictional composites rather than named organisations. It covers why reliability is the first AI investment, how temporary customer context keeps service AI helpful rather than intrusive, what public-sector procurement reform teaches any technology programme, and why digitalisation fails when it moves a broken process onto a screen. Strelitzia Bank and Vaalrand Metro are fictional, and every figure is illustrative.
4 articles
- Part 1Reliability comes before AI: why a bank's AI story often starts with an outageThe most convincing enterprise AI stories rarely begin with a model. They begin with a long outage, an uncomfortable re-platforming and the slow return of engineering time. Part 1 of South African Enterprise AI in Practice explains why reliability is the first AI investment.8 min read
- Part 2Helpful or intrusive? Temporary customer context in service AIService AI feels helpful when it knows why you are calling, and intrusive when it knows things you never expected it to. Part 2 explains temporary customer context: assembled for one conversation, used, then discarded, and why that boundary matters more than the model.7 min read
- Part 3What public-sector procurement reform teaches technology programmesBuying an eProcurement system does not reform procurement. Part 3 follows a fictional metro through a procurement reform that started with process, demand planning and shared data definitions, and treated the platform as the fourth decision, not the first.7 min read
- Part 4Digitalisation is not moving a broken process onto a screenA paper process with nine steps becomes a digital process with nine steps, and everyone wonders why nothing improved. Part 4 explains why digitalisation has to redesign the work, the roles, the data and the governance together, and where AI makes the old mistake faster.6 min read
Working through any of this in your own organisation?
These articles come out of delivery work rather than theory. If the problems in this series look familiar, a short scoping conversation is usually the quickest way to work out what is worth doing first.
Talk to CloudNala