Gartner Data & Analytics Summit Sydney 2024: AI and Data Governance

The Gartner D&A world tour continues! We were thrilled to join friends, customers, and colleagues at the latest conference. The Gartner Data & Analytics Summit Sydney 2024 delivered fresh insights into how leading APAC organisations are tackling the thorniest challenges in data management today, with a special focus on putting AI into practice for the business. In this blog, we’ll share our key takeaways from an outstanding keynote presentation and thought-leader panel featuring data innovators at some of our most forward-thinking customer companies to date. Let’s dive in!

Keynote: Collective Intelligence as a New Paradigm for Data and AI

Gartner’s Rita Sallam, VP Analyst and Mark Beyer, VP of Research, led a keynote with profound insights into the evolving landscape of data and AI. The central theme of the discussion was "collective intelligence," where the combination of human and machine intelligence enables organisations to solve complex problems more effectively. "Collective intelligence is how we solve problems,” Sallam said. “It's how we as humans and now humans and machines get things done together."

Sallam expanded on this by highlighting how the public sector is embracing collective intelligence. She cited Auckland Transport's innovative use of a generative AI conversational interface to streamline responses to public inquiries, showing how combining data, AI, and human input can lead to impactful solutions.

Beyer and Sallam expanded on the animal metaphors, likening AI to a "thundering stampede of wild horses"—chaotic yet full of magnificent possibilities. This aptly captures the essence of AI in today's world: exciting, fast-paced, and, if not managed correctly, potentially harmful. The dialogue underscored the need for a strategic AI approach, emphasising the importance of making data "AI-ready" and executing projects effectively to drive value creation.

Indeed, generative AI has advanced even more quickly than a band of wild horses, as the speakers touched on its role in democratising data and analytics. Beyer noted, "In only 23 months since the launch of ChatGPT, we've seen organisations making enormous progress in terms of preparing their data and their workforce." This rapid adoption underscores the need for organisations to keep pace with technological advancements and leverage them for competitive advantage.

The speakers also pointed to the octopus, with its remarkable balance of centralised control and distributed intelligence, as a powerful metaphor for operationalizing data management. Just as an octopus uses its arms autonomously while acting cohesively, organisations must empower decentralised teams. This approach enables agile decision-making and maximises the collective intelligence of the workforce.

As AI technologies evolve, so too must the skills and mindset of the workforce. The dialogue highlighted the importance of data literacy and the need to extend it to AI literacy. AI is not just the domain of data professionals; it affects everyone in the organisation. Therefore, educating leadership and building a data and AI-literate workforce is crucial. This effort is essential for realising the business value of AI investments and fostering a culture of collective intelligence.

Sallam concluded with a call to action, encouraging leaders to focus on execution and governance, linking data initiatives to tangible business outcomes. "Governance is a way of improving execution, not a barrier as often the business thinks," she asserted. The emphasis was clear: effective data and AI governance are crucial for creating value and should be communicated as such to leadership.

Data Leaders Panel: 3 Steps to Data and AI Readiness

On a (slightly) smaller stage, a panel of industry leaders convened, along with myself acting as moderator. Three data leaders from across financial services, telecommunications, and insurance, shared invaluable insights on the evolving landscape of data governance and quality in the age of AI. This discussion revealed how leading companies are shifting their strategies to better manage data scale metadata management, and enhance data quality to support real business outcomes.

From Governance to Enablement

One of the key takeaways from the panel was the evolving approach toward data governance. One leader revealed how his organisation positions data governance by avoiding the word altogether.

Instead of calling their function a data governance function, the organisation refers to it as a data enablement function. This shift in terminology is driven by the understanding that governance can sound restrictive and make people apprehensive, while data enablement focuses on helping people do their jobs better, delivering value to them, and making a difference with data. This rebranding reflects a broader industry trend toward framing data governance in a more positive, business-friendly light, emphasising support and facilitation over control and restriction.

I supported this view, suggesting that data governance should be an embedded part of job descriptions rather than an additional duty. This integration helps mitigate resistance and fosters a culture where data governance becomes a natural part of daily operations. By embedding data governance within existing workflows, organisations can reduce the friction of the perception that you’re adding to peoples’ workloads, making it easier for employees to engage and comply.

Scaling Metadata Management

As data volumes grow exponentially, the challenge of scaling metadata management becomes more pressing. One panellist discussed the complexities involved, highlighting that scaling up metadata management is often overlooked but crucial for maximising data value. Effective metadata management should be treated as a strategic initiative embedded within the broader data strategy. This includes developing and executing a roadmap to address scalability issues effectively.

This sentiment was echoed, emphasising the value of a federated approach for metadata management. By integrating it as an enablement layer within the overall strategy, companies can leverage existing organisational structures and embed metadata management as an implicit expectation in various roles. This approach minimises resistance and enhances stakeholder buy-in by aligning metadata management with existing responsibilities, while clearly demonstrating the direct benefits to individual business units.

The Crucial Role of Data Quality

Data quality emerged as a pivotal theme in the discussion, particularly regarding AI initiatives. The panellists unanimously agreed that high-quality data is essential for accurate and reliable outcomes in AI and machine learning models. One leader highlighted that despite over 25 years in the data industry, data quality issues persist and may even be worsening; the business ownership of data is critical, as technology alone cannot solve these issues. To address this challenge, some companies have implemented operational scorecards to track data quality, holding business teams accountable for their data inputs.

Creating transparency around data quality issues is vital. Data quality still remains their biggest challenge, and it's crucial to clearly articulate, identify, discuss, and then prioritise what needs to be fixed. This transparency allows organisations to tackle data quality issues strategically, focusing on the most impactful areas. Tools like Alation provide visibility and enable better data governance practices, which in turn improve the quality and usability of data.

Strategic Alignment and Data Governance in AI

As organisations increasingly adopt AI, aligning data governance with AI strategy becomes critical. One leader emphasised that a solid data strategy is essential for AI initiatives, as the reliability of AI models depends on data quality. He advocated for operationalising data governance through automation and tools to manage data quality and ensure consistent application across the organisation.

When leveraging generative AI, it's crucial to ensure all inputs are relevant, high quality, and contextually understood. A robust data governance framework enhances efficiency in traditional AI by reducing the time spent on data wrangling. With proper governance, the quality, origin, documentation, and context of the data are well understood, streamlining the entire AI process.

How are these organisations putting governance into practice? One panellist shared that implementing Alation was a “game changer” because a lot of these data processes already existed implicitly. With the data intelligence platform, his team could make obligations around data quality explicit and transparent while also making these tasks easier. Launching Alation really “made a big difference in getting buy-in and adoption.”

Parting Words

I concluded the panel by challenging the three leaders to share three key lessons they’ve learned from their respective data culture journeys. They shared:

• Data is a team sport, and it’s vital to recognise individuals for their contributions to data work by formalising the rules. 

• Implement a data literacy program to ensure the organisation understands data and its potential with AI, and to gain recognition for these efforts. 

• Utilize technology to unify the organisation, fostering a shared understanding and a common language around data.

• Leverage existing organisational structures when implementing a governance framework, rather than reinventing the wheel. 

• Transparency and communication are crucial in breaking down data silos, which persist without proper communication. 

• Establish a centralised source of truth to enhance transparency and communication. 

• Instil the concept of ownership in the source of data is vital for effective governance.

Conclusion

The panel's discussion underscored the evolving nature of data governance, from a restrictive, legacy approach to one of empowerment. Indeed, the shift towards a more enabling and integrated approach, the emphasis on scalable metadata management, and the critical importance of data quality all reflect the growing recognition that data governance is not just a regulatory requirement but a strategic asset. As organisations continue to navigate the complexities of data management in the digital age, these insights provide a valuable roadmap for harnessing the full potential of their data assets.

Today, data and analytics leaders are at the forefront of a transformative era. They have the opportunity to shape the future of their organisations and industries. This is a time to think big, execute well, and harness the collective intelligence of teams. The journey may be challenging, but the rewards—innovative solutions, enhanced business value, and a positive societal impact—are worth the effort.