AI automation
Most AI projects do not fail on the technology.
They fail because nobody diagnosed the process first, nobody governed the data, and nobody prepared the people who would have to live with the result. So we do those things — in that order — and we do not start building until we can show you what it is worth.
AI automation is not about replacing your people. It is about removing the repetitive, error-prone work so they can do the work that needs judgement.
Method
ADAPT™ — five phases, in order.
You cannot Transform before you Assess. The order is the method.
Assess
Process discovery, AI readiness, and an honest baseline you can measure against later.
Design
A prioritised roadmap with indicative ROI. The things worth automating, and the things that are not.
Architect
Solution design, integration points, data flows, and the governance around them.
Pilot
Build one thing. Measure it against the baseline. Prove it before you scale it.
Transform
Roll out, monitor, optimise, and hand the capability to your own people.
Investment
Three tiers. Start at the diagnostic.
All figures in South African Rand, excluding VAT. Third-party platform and API costs are passed through at cost — we do not mark them up, and we tell you what they are before you commit.
AI Opportunity Assessment
A focused diagnostic. You leave with a complete process inventory, an AI readiness assessment, and a prioritised roadmap with the numbers attached.
- Discovery workshops and interviews
- Full process mapping
- Prioritised opportunity roadmap
- Indicative ROI per opportunity
Pilot Implementation
Design and build of one priority initiative — a workflow automation, a document-extraction pipeline, an AI agent — deployed, measured, and proven.
- Solution design and architecture
- Build and pilot deployment
- Measurement against the baseline
- Change management and training
- 90-day post-go-live warranty
Scale & Managed Service
Organisation-wide rollout, governance, monitoring and continuous improvement once the pilot has earned it.
- Rollout across the organisation
- Managed AI operations
- ROI dashboards and quarterly reviews
- Continuous optimisation
Why the diagnostic is not optional. Tier 2 pricing is a range because the honest answer depends on process complexity and how many systems have to be integrated. Anyone who quotes you a firm number before assessing your processes is guessing, and you will pay for the guess later.
Governance
The part most AI vendors skip.
Automation that makes decisions about people is automation that will eventually be audited. We build for that day from the start.
- Human-in-the-loop on every high-risk or high-value decision. AI augments; it does not act unchecked.
- Audit logging of automated decisions and actions.
- Model and prompt version control, with monitoring for accuracy and bias.
- Data sovereignty — where personal data cannot go to a third-party AI API, it does not.
- Escalation and fallback for when a model is uncertain or fails.
Before we touch your data
Signed, not assumed.
- Mutual NDA and a written Data Processing Agreement, in place first.
- Role-based, least-privilege access across every tool.
- Encryption in transit and at rest; secrets handled properly.
- Defined retention, residency and deletion — including erasure requests.
- Full transparency on every sub-processor and AI provider in your solution.
Find out what is worth automating.
The diagnostic is fixed-fee and it stands on its own. Even if you never build anything with us, you leave knowing which of your processes are worth the money and which are not.