Sonia Boije, Chief Data Officer at nbn, shares how simplification and governance are creating trusted data foundations for AI and enterprise growth.
In this interview with ADAPT’s Principal Research Analyst Peter Hind, Sonia outlines how the organisation is reshaping its approach to data.
Through simplification, governance, and continuous improvement, Sonia is positioning nbn’s data as both a trusted business asset and a foundation for AI innovation.
Simplifying the data landscape
Sonia stresses that digital transformation begins with solid foundations.
For nbn, that meant consolidating multiple, fragmented data discovery environments into a single, modernised platform.
The shift, driven by both IT and the business, was framed as an 18-month program with a clear deadline for retiring legacy systems.
This created urgency and collective ownership across teams.
Simplification was not only about efficiency.
It gave staff access to cutting-edge technology: better data lineage, higher data quality, and a more consistent user experience.
But Sonia acknowledges that getting business users on board required patience.
Initially sceptical, they eventually saw the value once they understood how modern platforms could accelerate workflows and unlock new opportunities.
Accelerating with AI while upgrading legacy systems
While simplifying the data estate, Sonia also experimented with generative AI to increase productivity.
One example was using AI to help rewrite and generate SQL code for data quality checks, with humans in the loop for validation.
This approach not only accelerated delivery but also demonstrated how emerging technologies could support legacy simplification.
Redefining governance through data products
A central pillar of nbn’s strategy is shifting from abstract data governance to tangible data products.
Sonia uses the analogy of a cereal box: a data product should have a brand people trust, clear “nutritional information” like quality thresholds, lineage, and access rules, and even an expiry date.
Like consumer goods, data products must be continuously enhanced to remain valuable.
This model fosters stronger producer–consumer relationships.
Producers gain visibility into who uses their data and for what purpose, while consumers can rely on defined service levels.
For Sonia, this creates accountability, transparency, and a more dynamic data ecosystem that can also feed advanced AI models.
Making data quality an organisation-wide priority
Sonia has elevated data quality to an executive priority under nbn’s 2030 vision.
Every employee is expected to play a role, from flagging process issues to resolving technical gaps, while domain managers remain accountable.
Progress is tracked at the board level, underscoring its strategic importance.
To classify data effectively, nbn uses a framework that includes categories such as commercial sensitivity and privacy.
Recognising the manual burden, the team is automating classification with generative AI, again keeping humans in the loop for risk management.
Policy-based access controls allow faster, compliant use of classified data across the organisation.
Measuring value beyond commercial outcomes
Sonia emphasises that the value of data cannot be measured only in financial terms.
Data helps reduce risk, protect brand reputation, and ensure regulatory compliance.
She cites the example of marketing communications, where accurate location and service data are crucial to avoid penalties or customer dissatisfaction.
Continuous improvement and cultural momentum
For Sonia, data transformation is a journey without a finish line.
Threats evolve, technologies advance, and strategies must adapt.
To keep momentum, nbn has embedded rituals like two-week sprints and end-of-sprint demos.
These showcase progress, celebrate contributions, and maintain energy across the team.
Continuous recognition, she says, is key to sustaining commitment in what is otherwise a long-term, relentless journey
Key takeaways
- Commit to simplification: Set a clear timeline to retire legacy systems and unify fragmented environments.
- Turn governance into products: Package data with clear ownership, quality metrics, and lifecycle management.
- Make data quality cultural: Elevate it to an executive priority and embed responsibility across the organisation.
- Use AI pragmatically: Apply generative AI to accelerate classification, quality checks, and legacy upgrades while keeping humans in the loop.