In summary:
- AI and cloud accelerate outcomes but expose data weaknesses, turning poor quality, silos and unclear ownership into visible blockers rather than hidden issues.
- Strong data foundations determine whether AI creates value or chaos, shaping trust, compliance, explainability and decision‑making at scale.
- AI readiness goes beyond technology, requiring robust data strategy, governance, ownership and literacy to enable safe, repeatable innovation.
As organisations accelerate their adoption of AI and cloud technologies, many expect clarity, efficiency and smarter decision-making. Yet in practice, faster technology often feels like greater complexity.
In our UKOUG webinar, James Herbert, Oracle Business Development Director at Sopra Steria, and Susannah Matschke, Head of Data and AI Foundations, explored why this happens and why strong data foundations are now the defining factor between AI success and AI-driven chaos.
Acceleration doesn’t create problems - it exposes them
AI and cloud dramatically increase speeds of deployment, analytics and decision-making. But that acceleration also removes the buffers that previously masked underlying issues.
Poor data quality, fragmented ownership, inconsistent definitions and siloed systems don’t disappear in an AI-enabled world, they become impossible to ignore.
AI doesn’t introduce these challenges. It amplifies the weaknesses that already exist.
When foundations are strong, AI accelerates progress. When they are weak, AI accelerates confusion.
Oracle AI: embedded across the stack
Oracle has embedded AI across every layer of its technology stack, from infrastructure and applications through to analytics and enterprise performance management.
The key capabilities include:
- Embedded AI in Oracle Fusion applications, available as standard.
- Machine learning within EPM, supporting forecasting, budgeting and variance analysis.
- Oracle Analytics Cloud, enabling predictive insights across multiple data sources.
- The Oracle AI Data Platform, acting as the semantic and integration layer across the estate.
Crucially, Oracle’s approach allows organisations to bring their preferred large language models such as OpenAI, Cohere, Gemini or others, while retaining control over security, governance and data residency.
Even the most advanced technology can only perform as well as the data that fuels it.
Data Is the fuel for AI and quality matters
Data is often described as the fuel for AI. But feeding poor-quality, inconsistent or unclear data into advanced systems is like putting the wrong fuel into a high-performance engine.
No matter how sophisticated the tooling, weak data foundations lead to poor decision-making, compliance and regulatory risk and limited trust in analytics and AI outputs, while increasing levels of rework.
In public sector and regulated environments especially – where Sopra Steria delivers many large contracts - confidence in data lineage, quality and explainability is non-negotiable.
AI readiness Is bigger than technology
Many organisations frame AI readiness as a tooling or skills problem. In reality, it is far broader.
Sopra Steria’s six-pillar AI readiness model spans:
- Technology and infrastructure.
- Governance and ethics.
- Skills and capability.
- Operating model.
- Culture and ways of working.
- Data foundations at the core.
Data underpins every pillar. Without trusted, consistent and accessible data, none of the others can operate effectively.
The real causes of AI chaos
Across sectors, the same data challenges appear consistently. Data silos preventing a full view of the organisation, so that different teams getting different answers to the same question – this undermines trust.
Manual reconciliations dominating analysts’ time means less time spent analysing what the data is telling the business. A lack of clear ownership and accountability, combined with governance that either blocks progress or introduces risk.
These problems existed before AI. The difference now is speed and the cost of getting things wrong.
What a data strategy really is (and isn’t)
A data strategy is not a document, a slide deck or a one-off exercise.
A real data strategy answers practical, operational questions:
- Which data matters most to the organisation?
- Who is accountable for it?
- What decisions depend on it?
- What risks must be managed?
- What outcomes are we optimising for?
The ultimate test is simple: if the strategy doesn’t influence day-to-day decision-making, it isn’t working.
Foundations turn strategy into reality
Data foundations are what make strategy tangible and executable. They include:
- Clear ownership and stewardship.
- Defined standards and shared definitions.
- Visible data lineage and quality controls.
- Proportionate governance that enables, not blocks.
- Accessible data with shared understanding.
- Investment in data literacy across the organisation.
Value doesn’t emerge automatically just because data exists. It requires structure, care, attention and ongoing stewardship.
Building confidence instead of chaos
When data strategy and foundations align:
- People trust the numbers.
- Decisions happen faster, with less debate.
- AI initiatives scale safely and repeatably.
- Innovation accelerates without increasing risk.
Strong foundations aren’t the opposite of innovation, they are what make innovation reliable.
If you’d like to discuss your organisation’s data and AI foundations in more detail, get in touch with Susannah Matschke, Head of Data and AI Foundations at Sopra Steria, to explore how to build trusted, scalable AI readiness.