CJ Nakamura (CJ): To start, I wanted to discuss your recent article in CDO Magazine on the Digital Data Steward. What was the core inspiration behind developing the DDS concept? Furthermore, for our practitioners, what is the Digital Data Steward actually, and what key takeaways should they focus on regarding this agentic system, and what is your call to action on this initiative?

 

Maria Villar (Maria): The core inspiration originated from years of observing a persistent, painful, and critical challenge in the CDO ecosystem: the persistent challenge to effectively staff and setup the traditional Data Steward leadership role. This role is pivotal - it is the bridge between the business and the data  - but it requires a rare blend of technology knowledge, data knowledge, deep business process expertise, and strong leadership, making it incredibly difficult to staff effectively. Additionally, the few people who have these roles in the business teams are busy doing other important business functions.

 

The Digital Data Steward (DDS) is the solution born out of a collaboration of myself and three colleagues who I asked to contribute to the ideas: Mike Alvarez, Beth Hiatt and Christine Legner. It is an agentic system, meaning it involves multiple, specialized AI agents working together with robust orchestration, designed to automate the majority of the Data Steward's tedious, day-to-day tasks. The viability of the DDS is entirely due to the maturity of generative AI and agentic systems today; this simply wasn't possible five years ago. This system is designed to be always on, always listening, and always adapting to the environment in a way a human cannot.

 

By automating the "doing role," the DDS transforms the human Data Steward into a strategic leadership and oversight role, managing the agentic system and focusing their energy on high-value business alignment and strategy. We defined four key agents that operate in concert: a Data Strategy Agent, a Data Quality Agent, a Metadata Agent, and a Retention Agent. We’ve demonstrated this concept with a "DDS Cockpit" visualization to show CDOs how this automated governance system would operate with real code.

 

The Call to Action: For the DDS to evolve into the reliable, robust, and scalable solution we need, it cannot be built in isolation. This is an open invitation. I ask the global data community, alongside innovative partners like AI81Works, to actively engage with this model. We need people to contribute their unique, specialized experience, subject matter knowledge, and deep skills to further improve and "train" this big DDS model. Collaboration is essential to ensure the Digital Data Steward truly solves the decades-old staffing crisis and ushers in a new era of proactive, automated data governance.

 

CJ: Reflecting on your extensive career as a senior data executive with SAP, Fannie Mae, and IBM, what core principles have consistently guided your approach in leading any of these transformations?

 

Maria: I rely on three core principles that have proven essential for navigating any complex data transformation:

First off - Business First, Data Second (Be a Business Leader): This is non-negotiable. A successful data executive must be viewed by the organization as a business person first. Your credibility must be rooted in business value and making reasonable business trade-offs.

Secondly - Resiliency is Key (The "Not Yet" Mentality): My personal mantra was, "No”

was never an answer. “No” was always “not yet". You must possess high resiliency, treating any initial "no" as a signal that you need to go back and build a stronger, more quantifiable business case.

Thirdly - Build a Coalition (You Need Lots of Friends): Data is inherently cross-functional. Success requires a broad, unified front of support and partnership across the entire C-suite and all major business peers. A CDO must be a master coalition builder.

 

CJ: The Chief Data Officer role has certainly evolved over the years. From your vantage point, how has the CDO function shifted - from initially being about defense and compliance, to a more offensive posture, and now towards full enterprise-wide value creation?

 

Maria: We have seen that clear evolution. The CDO role began firmly rooted in defense and compliance. It then successfully pivoted to the offensive side, focusing on data monetization. Today, the core mandate is truly about full enterprise-wide value creation. However, despite this evolution in mandate, the core organizational challenges have been dishearteningly consistent for over two decades. The expectations have changed dramatically, but the fundamental issues - like securing enterprise-wide buy-in and getting the necessary budget - have not improved.

I see many incoming CDOs simply not set up for success. They often spend their pivotal first 100 days doing damage control or having to explain the very purpose of their office, which quickly creates a negative perception that they aren't delivering value. The failure is systemic, not individual.

 

CJ: Looking at the broader landscape, what do you see as the most common barriers to scaling AI adoption across legacy enterprises? How have you personally approached overcoming those challenges?

 

Maria: The key barriers to scaling AI adoption are consistently organizational and cultural, not technical.

1. Cultural Inertia: Organizations are resistant to change, and AI is often wrongly perceived as solely an "IT problem," leading to siloed adoption.

2. Resource and Expectations Gap: The mandate for value creation far outstrips the budget provided, leading to 'death by pilot project.' You cannot scale without the necessary resources.

3. Governance Paralysis: Rigid, old data governance models are mistakenly applied to new, agile AI initiatives, which chokes innovation and slows adoption speed.

4. The Data Foundation Gap:  the data needed for the AI use cases is not available or not reliable, not built on a solid foundation for use

My approach has always been rooted in Pragmatism and Persistence. First, Pragmatism dictates an ROI mindset. You must focus on the measurable value first to shift the conversation from a cost center to a value generator. Second, Persistence is tied to resiliency - you pivot, find a stronger champion, and demonstrate value in a small, irrefutable way. Finally, we must apply "Smart" Governance to overcome paralysis - governance that is automated, embedded, and an enabling service.

 

CJ: Hiring and developing the right data and AI leaders is obviously critical to success. From your experience, what are some of the biggest mistakes organizations commonly make here? What advice would you give to boards or CEOs who are trying to build their leadership capabilities in this space?

 

Maria: The hiring process is a common point of failure. The mistakes are critical:

Firstly - Failing to Vet the Role, Not Just the Person: They hire a brilliant candidate before defining the purpose, scope, and required authority of the role across the enterprise. The responsibilities and budget must be agreed upon by the CEO, CIO, and business leads before the CDO starts.

Secondly - The "Magic Wand" Fallacy: They hire an executive, give them no dedicated budget, and expect them to 'tin cup' their way to funding. If you don't provide the initial capital and resources, you have effectively set them up for failure from day one.

 

My Advice to Boards and CEOs: The Chief Data & AI Officer role is the most complex executive hire today. It demands a rare blend of technical depth, business acumen, and political savvyness. The old playbook of traditional executive search simply won't work for this niche. You must seek out specialized expertise. Working with highly specialized, data & AI-focused executive search and recruitment firms, such as AI81Works, is not an option; it is a necessity to avoid the costly mistake of a failed CDO hire and to ensure you land a transformational leader.

 

CJ: That being the case, and to close out today’s interview, what is your most important advice for the next generation of CDOs, particularly as they navigate the rapid acceleration of generative AI?

 

Maria: My final advice is focused on three keys for value creation in this new AI-driven era:

Firstly - Lead with an ROI Mindset: The new CDO must lead with an ROI-first mindset. Before proposing any strategy or project, you must first quantify the measurable value and benefit to the organization.

Secondly - Implement “Smart” Governance: Governance is necessary, but it must be done in a "smart" way. It needs to be flexible,  reliable, automated, embedded, and always-on. Move away from rigid standards to an approach that is reasonable and enables safe innovation at speed.

Thirdly - Recognize and Embrace "Shadow AI": Leaders must acknowledge and enable "Shadow AI." This refers to the organic, bottom-up use of LLMs (like ChatGPT or Gemini) by employees to boost personal productivity. This is the fastest way AI will be adopted. Leaders must find a way to quantify and enable that value, recognizing that this is an acceptable, pragmatic path to AI introduction and transformation.

 

CJ: Maria, this has been an incredibly insightful and practical conversation, offering clarity on everything from the Digital Data Steward to executive hiring best practices. Thank you for sharing your experience and your vision for the future of data leadership. We also echo your Call to Action: we encourage the entire data community to engage in improving the DDS model to secure the future of data.

​