IPT urges businesses to fix their data foundation before scaling AI
South African businesses are moving quickly from having an interest in AI to wanting the technology implemented. However, many are still trying to build new capabilities on data environments that were not designed for intelligent automation, analytics, or AI-assisted decision-making.
IPT, a managed IT services, cybersecurity, and AI consulting provider, says businesses need to get their data foundation right before investing in advanced AI use cases.
The issue is not whether AI can add value. It is whether the business has the data quality, classification, governance, and security controls needed to support it. Documents may sit across disconnected systems, sensitive information may be poorly classified, access rights may not reflect current roles, and data needed for reporting or automation may be duplicated, incomplete, or difficult to trust.
That becomes a practical problem when businesses start introducing AI tools, dashboards, process automation, and agentic AI use cases. AI systems depend on the information they are given. If the underlying data is messy, badly governed, or exposed to the wrong users, the business carries those weaknesses into its AI environment.
Nicol Germishuis, vCIO at IPT, says data readiness has become one of the most important conversations in AI adoption.
“Many businesses want to move quickly into AI, but AI does not fix poor data discipline. If information is duplicated, poorly classified, unstructured, or not properly governed, those weaknesses are carried into the business’s AI environment. The first step is to make the data usable, secure, and ready for the decisions or workflows AI is expected to support.”
IPT’s approach focuses on helping businesses prepare their data layer before scaling AI. This includes assessing data quality, reviewing process maturity, improving classification, strengthening governance, and helping organisations understand where Microsoft technologies can support a more controlled AI environment.
Microsoft Fabric’s OneLake is a single, unified logical data lake for an organisation, designed as a central place for analytics data. Microsoft Purview supports data governance, data security, and compliance capabilities, while Purview Information Protection includes data classification, sensitivity labels, and encryption to help secure information across Microsoft 365 and other environments. Microsoft Purview DLP policies help organisations identify, monitor, and protect sensitive information.
For businesses using Microsoft environments, the data layer is critical to AI readiness. Before AI can support better decision-making, automate workflows, or assist users with more advanced agentic models, organisations need to know where their data sits, who can access it, how it is classified, and what controls protect it.
IPT says this is where AI consulting must move beyond tool selection. The real work sits in aligning data architecture, security controls, user adoption, governance, and operational goals.
“AI readiness is not only about choosing the right platform. It is about understanding whether the business has the foundation to use AI responsibly and effectively. That means looking at data quality, governance, access, classification, and the processes that AI is expected to improve,” adds Germishuis.
IPT’s AI services include readiness assessments, AI consulting, AI-driven data analytics, business intelligence, adoption support, and managed intelligence operations. The aim is to help businesses move from fragmented data and isolated AI experiments toward more structured, secure, and measurable AI adoption.
Germishuis says businesses should treat AI as an operating discipline, not a once-off deployment.
“AI deployed is not AI done. The value depends on the foundation beneath it and the way the environment is managed over time. Businesses that get their data layer right will be in a stronger position to use AI in ways that are practical, secure, and aligned to how they actually work.”
IPT’s AI services include readiness assessments, AI consulting, AI-driven data analytics, business intelligence, adoption support, and managed intelligence operations, helping clients connect AI ambition to the underlying data and governance work.
