Interviews
Kimberly Nevala Director of Business Strategies at SAS – Interview Series
Kimberly Nevala is the Director of Business Strategies for SAS Best Practices, where she advises on the strategic value and practical impact of emerging analytic applications and information trends. Her expertise spans business analytics, data strategy, governance, business-IT alignment, and building analytic cultures.
Previously, Kimberly was a Principal Consultant at Baseline Consulting Group, a leading management consultancy specializing in business and data strategy. There, she guided organizations worldwide on BI and analytics, data governance, and business transformation. Drawing on 20 years of hands-on experience and forward-looking insights, Kimberly helps organizations unlock the full potential of their data.
SAS is a global analytics and AI company offering solutions in data management, advanced analytics, and artificial intelligence. Its flagship platform, SAS Viya, emphasizes speed, scalability, and trusted decision-making for organizations across industries. The company showcases use cases in fraud detection, marketing optimization, IoT, and risk management, while also providing free trials, training, and academic programs to support widespread adoption of analytics.
You’ve built your career guiding organizations through complex cultural, strategic, and technical transformations—from management consulting to leading business strategy at SAS. How has that journey shaped your view on what it truly takes for enterprises to adopt AI in a meaningful—and inclusive—way?
Despite the expansive scope of emerging AI capabilities and their equally seismic ramifications, the basic building blocks for driving technical transformation have not changed. Although the ramifications for getting it wrong have.
First, there needs to be a clear articulation of the organization’s values and priorities. Along with a clear understanding of how these factors influence how employees or customers engage with the brand. Next, there needs to be a broad understanding of the emerging tech, including its innate limitations and constraints. Finally, visibility into foreseeable skill gaps and barriers to adoption based on one and two above. Notably, this is not limited to technical expertise but also business acumen.
All of which must then be supported by robust but adaptive governance. Underpinned, of course, by targeted, deliberate investment in both people and the technology.
You’ve spoken about AI literacy rising to the level of essential leadership skills like strategic thinking and financial acumen. What does AI fluency look like in practice for today’s leaders?
The litmus test for practical fluency is straightforward. Do your leaders have the knowledge required to intelligently interrogate a proposed AI solution?
This is not to suggest leaders need to get finger-deep into code. Rather, leaders at all levels should demonstrate contextual competency. Meaning they are educated in the factors that could impact or sway their judgment of an AI system. These factors will change based on the leader’s role. Specifically, the scope of the decision they are charged with making.
Put simply, the test is this: do they know what questions to ask? For example, can an executive identify the human and technical factors with the greatest impact on the viability and risk profile of a given AI solution? Both broadly and in an applied context.
Even better: do they understand the limitations of the technology? PR and marketing tends to focus on the upside, meanwhile experiential training exposes limitations. These limitations can dictate the procedural and technical guardrails required to deploy a given AI solution gainfully. In this regard, the extent to which leaders understand and identify such limitations is often the difference between meaningful adoption and failure.
Why is gender equity in AI leadership such a pivotal issue right now—and what risks do we face if women remain passive adopters rather than active shapers of AI’s development and governance?
I am privileged to work, speak with, and learn from women working across all aspects of AI every day. So, the framing of women as passive adopters, for me, evokes images that are childlike and misleading. Women are actively involved in leading AI efforts, albeit in smaller numbers than our educated population would indicate.
Gender inequity in tech is a multifaceted issue that predates AI. There is no single solution as there are numerous factors at play. Some result in less funding for and/or recognition of women in the field. Making it painfully ironic that women who do receive broad recognition are often working in responsible or ethical AI. Initially labeled as anti-AI, their work is now being reconsidered as attitudes evolve.
Regardless, without deliberate intervention, we risk AI systems that exacerbate existing inequities at scale. We see this in the use of automated HR screening and hiring systems that prefer male candidates due to historical biases captured in the data.
A lack of diversity at all levels means that decisions regarding what types of AI systems are built, how they are deployed, whose perspectives and needs they address and whose they don’t, are dictated by an increasingly insular group. This is bad for everyone. Diverse management teams drive better business results. Diverse product teams deliver AI solutions that are more robust, resilient, and work better.
In the end, if AI systems don’t address the perspectives and needs of women (or any other underserved demographic) or are shown to be actively harmful, what is the incentive for adoption? Except, of course, that these systems will be deployed with or without our participation. By including diverse voices and enforcing AI governance it is possible to deliver on AI’s promise.
Given SAS’s latest innovations—such as AI agents within SAS Viya, intelligent decisioning frameworks, synthetic data tooling like Data Maker, and custom AI models—how are enterprises rethinking their leadership pipelines to elevate both AI fluency and equity?
Enterprises are realizing that the scope and impact of these applications outstrip the capacity of any single decision maker or doer. Deploying advanced AI successfully depends on the trust and engagement of the “village.” The potential issues and gotchas are far too wide for any leader to effectively address alone.
In recognition of this reality, leading AI enterprises are leaning into participatory design and governance practices. In this model, fluency and equity are not enhanced by formal literacy and training programs nor by HR initiatives alone. But by the knowledge sharing and collective decision making that occurs when bringing diverse perspectives from across the enterprise ecosystem together. These collaborations are not limited strictly to decision makers but actively include the people whose work will be affected by the AI system under consideration. While primarily the domain of public and governmental agencies today, some organizations have actively engaged external stakeholders including customer groups and public interest organizations.
What practical steps can aspiring leaders—particularly women—take today to build credibility and agency in an increasingly AI-driven marketplace?
First is to recognize that we do have agency. Every individual has the opportunity to influence the development of AI in a number of ways: through the deliberate consumption and use of particular AI tools; through advocacy and collective engagement in community forums and local politics; by joining employee interest groups or community champion networks (and if those don’t exist, by starting one); and by raising your hand to participate in AI programs or lead initiatives at work.
Second, is to invest in your own literacy at whatever level makes sense to you. There are numerous online resources spanning from coursework on AI governance and ethics, to online college-level introductory classes covering the full spectrum of AI capabilities. Many of these options are provided for free or for nominal charges.
Finally, join and actively contribute to the AI industry or special interest networking groups. These provide an opportunity to not only share knowledge but curate relationships that can provide mentorship, alliances, and credibility as you build your expertise and your network.
You’ve long emphasized cultural transformation, data governance, and responsible adoption of technology. How do these foundational elements support the development of AIready leaders?
The traditional answer is that these foundations arm leaders with the tools to address complexity, manage risk, lower cost, and increase productivity. All of these outcomes are true, however, the most underappreciated benefit of these practices is confidence.
Done right, these elements create an environment in which employees can experiment and innovate confidently within the bounds of the company’s ethical, legal, and business ethos.
This is particularly important as available solutions run the gamut from personal productivity tools to engineered, embedded workflows. Decisions about AI use do not just happen in the boardroom or in the context of a formal AI project. From development to exploration to deployment and beyond, a myriad of compounding factors influences how the system performs and is perceived.
Responsive guidance enables individuals and teams at all levels to actively experiment, make timely decisions about if and how to advance solutions, and build better within the confines of a system of regulations.
This direct link between responsible, adaptive innovation and governance is no longer a hypothetical. IDC and SAS recently undertook research on the link between responsible AI practices, trust and value realization. Responsible innovation is not a theoretical good but increasingly an economic imperative.
From conversations on your Pondering AI podcast, what insights on responsible, inclusive leadership stand out as most critical for organizations navigating AI transformation?
The language of AI is tricky. We speak of AI systems that think, hallucinate, and even misbehave. We conceptualize AI systems as teammates or colleagues. As such, it’s tempting to transfer assumptions and expectations from humans to AI. Thereby we lose focus on human accountability, participation, and responsible use. We also obscure the rigor required to deliver robust, resilient AI systems which, when done well, appear simple.
This makes it critical to center our understanding of AI as an innately human endeavor. Despite prevalent anthropomorphic narratives, AI systems always reflect the values and intentions of their creators: the problems we choose to solve, the techniques or models we apply, the data we use in training, and how users engage with the system. Every AI system is the product of these human decisions.
Confidence and trust in AI systems does not come from AI. Humans ultimately decide to automate or augment a person’s job. It is not the system which determines what level of accuracy or explainability is sufficient. It is not AI that determines the most appropriate user interaction model. When a chatbot misbehaves, it doesn’t impact the AI; it degrades the brand and bottom line of the company wielding the AI.
Therefore, human-centered AI is key.
Many organizations have advanced AI toolsets—but still struggle with culture, governance, and inclusion. From your experience, what interventions or mindsets help organizations bridge that gap effectively?
AI, like many emerging technologies, is subject to falling into the same hype cycle as its predecessors. During the big data era, becoming data-driven was all the rage. Few organizations delivered on their vision. Those that did were disciplined about solving identifiable problems such as reducing fulfillment times for X or enabling self-service for Y. It is no coincidence that many of those same companies are now industry leaders in AI.
When identifying AI initiatives, start with clear business objectives and measurable outcomes, then identify which available AI techniques are fit for purpose. Resist the urge to be flashy. Innovative solutions to boring problems ultimately pay dividends. Address dull, repetitive processes and tasks that happen at scale.
You won’t get it right all the time. But focusing your attention and investment with the business outcome in mind allows for the highest probability of success.
As AI systems become more autonomous and embedded into decision-making—through tools like AI agents and copilot assistants—what ethical guardrails or governance structures do you believe are essential to uphold trust, transparency, and equity in leadership?
At SAS, we define AI governance through the lens of oversight, compliance, operations, and culture. Regardless of which framework you adopt, there are two factors I would underscore.
The first is that every AI application, even if built on the same underlying model or data sets, must be evaluated independently and on an ongoing basis. What works for marketing may not fly for finance. An acceptable hallucination, aka error, rate for employees will not be effective in customer interactions. In addition, due to their probabilistic nature, shifts in user behavior, underlying data flow, and a myriad of other factors can quickly change an AI system’s behavior and outputs.
The second is that AI ethics and governance do not exist in a vacuum. The nature of these solutions necessitates collaboration with corporate governance, legal and regulatory compliance, enterprise risk management, and cybersecurity. Leveraging and extending these practices rather than reinventing them can provide a running start.
Looking ahead, how do you see leadership competencies changing in an AI-driven world—and what are the implications for organizations that don’t cultivate that new fluency?
I believe foresight and resiliency will become increasingly important. The rapid pace of change encourages leaders to continuously assess and adapt business strategies and tactics as the technology evolves. Leaders must assess value, viability, and risk across different time horizons and perspectives. Given the sociotechnical nature of AI, those perspectives must increasingly include not only shareholders, but customers, employees and, at times, society at large. Foresight is required to navigate this landscape.
Resiliency is also a factor in AI. As Jordan Loewen-Colón eloquently notes in reference to AI adoption and governance: AI is a process not a switch. The ability to intentionally learn, adapt and continuously improve while staying the general course will differentiate leaders from the rest.
Thank you for the great interview, readers who wish to learn more should visit SAS.
