The first and foremost consideration is having a clear business case and defined requirements. We need to understand not just how these emerging technologies might be used, but why and by whom. Beyond the business case, it is critical to evaluate integration, considering how the technology interacts with existing platforms internally as well as with external systems where interface requirements may exist.
We also assess the maturity of the capabilities themselves. Once potential solutions are identified, we often start with proofs of concept or incubation projects to test and evaluate their fit.
Finally, technology adoption is not just about the tools; it is about the people. We need to gauge organisational readiness, the level of change required, and the skill development needed to successfully implement these technologies. Essentially, the assessment spans technology, integration, and human factors.
A business-led approach is crucial and decisions must be grounded in the organisation’s mission and requirements.
When considering modernisation, there is often a choice between incremental improvements or a larger “big bang” shift. There is no single answer and the right approach depends on organisational maturity, current tools, and systems.
Some core principles include designing for interoperability, use of standardised interfaces, zero trust with security and privacy embedded by design, and ensuring governance is integral from the outset. Importantly, delivering value quickly is key.
Trust starts with the data itself. For AI-driven insights to be reliable, the data feeding these systems must be trusted. This requires clear data policies, standards and governance at the centre. This is particularly important in organisations that operate in a federated manner.
Building trust in AI requires clear governance and assurance processes to ensure that the use of AI is aligned with organisational policy and has appropriate risk management. It also requires clear accountability.
A human-centred approach is critical. We need to understand the impact of AI on people, ensure the systems are explainable, and be able to justify the outcomes produced. Reliability, security, and proportionate controls are all essential to maintaining confidence in AI applications.
Good data integration at scale means having consistent, discoverable, and well‑governed data flows across domains. This is enabled by API‑driven interoperability, strong master data management, and metadata catalogues that make data assets easy to discover and use..
Balancing centralised governance and domain autonomy requires a clear policy, standards, and framework at the centre while allowing domain custodians to manage implementation locally.
Domains remain responsible for data quality, supported by centralised governance with visibility and metrics to prioritise effort.
This provides a balance of central governance to ensure trust and consistency, with domain autonomy that provides speed and agility.
Foundational data skills like engineering, architecture, and modelling remain critical, but the differentiator is often people skills. The ability to collaborate with stakeholders, a mindset of curiosity, and a culture focused on supporting the broader organisation are essential.
Planning, managing, and delivering outcomes are equally important. Technical skills alone will only take you so far; success depends on the combination of technical proficiency, stakeholder engagement, and problem-solving curiosity.
Paul Robards is a speaker at CDAO Sydney. Interested in learning more about Data? Join us at CDAO Sydney this March!
Also don't miss Data & AI Architecture Sydney and Enterprise AI Sydney, happening in the same space!