CIO Masterclass: ‘Becoming a Scalable AI-Driven Enterprise’

Artificial intelligence (AI) is a captivating subject that resonates widely. While experimenting with ChatGPT and image generation can be thrilling personally and professionally, translating AI into large-scale business value proves to be more challenging. The recent CIO Masterclass by Anderson MacGyver provided actionable insights in this area, with contributions from Management Consultant Anton Bubberman and Frank Ferro, who spent the past decade overseeing Analytics, Data Insights, and GenAI at PostNL.

An informal survey among the attendees revealed that everyone had experience with ChatGPT. When moderator Fiep Warmendam asked participants to share their last-used prompt with the person next to them, it became clear that the tool is primarily useful for personal tasks—such as planning an exciting holiday destination, complete with the best routes, or selecting a new phone or other potential purchases.

Smartwatches, on the other hand, appeared to be less popular. Warmendam confessed she avoids letting her running routines be dominated by an abundance of data: “This likely influences your behavior and decisions. I fear it could take the joy out of running—I don’t want to lose the human touch.” This risk can also apply in business. However, data and intelligence can add value in other areas, like predicting delivery times for meals, groceries, or packages.

This set the stage for the insights and experiences shared by the two specialists. “In line with Roy Amara’s Law, we tend to overestimate the short-term impact of AI while underestimating its long-term effects,” noted Anton Bubberman. The senior Management Consultant is also Guild Lead Data to Value at Anderson MacGyver. Has extensive relevant data experience in sectors ranging from healthcare to energy and finance.”

Cognitive Skills

Under the ironically yet compelling title “Create a clickbait title for my AI-vision talk”, Bubberman introduced the concept of AI, which becomes more powerful as autonomy and adaptability increase. Ultimately, we are moving toward artificial general intelligence (AGI), which matches human cognitive skills. This would allow AI to independently perform complex tasks across diverse domains and adapt to new situations. However, since we are far from the AGI phase, human oversight and monitoring of AI remain essential.

Bubberman outlined three success factors for scalable and potentially value-creating AI deployment within organizations, using analogies from chess, jazz, and philosophy.

Chess is all about planning, foresight, and continuous evaluation. “Circumstances and opportunities are constantly evolving, and organizations and leaders must adapt. You must always think ahead to the next move on the chessboard.”

The connection with jazz is that while playing music and improvising might appear effortless, it often follows a long period of practice. Beyond technical skills, it requires an understanding of theoretical frameworks and foundational principles. “Dedication is necessary to master an instrument. It involves hard skills but also soft skills, such as interacting with other band members.” In the digital domain, a culture driven by AI is essential, alongside technical prerequisites, with attention to ethical considerations.

Finally, philosophy highlights the dual-edged nature of tools. A surgeon’s scalpel can perform miracles but, in the hands of an unskilled or malicious individual, it can cause disaster. Similarly, AI carries risks such as polarization, information bubbles, misinformation, and bias—particularly when data is incorrect or human oversight fails to address potential negative impacts. “In the right hands, AI has the power to positively change the world,” Bubberman concluded.

Lessons Learned

Frank Ferro reflected on his decade of experience in realizing business value with data and AI. He began his presentation with a cloud of personal data—trivia and relevant details that only gained meaning after verbal explanation. From his birth to the year 2025, when after nearly 17 years at PostNL and a temporary role as Program Director GenAI at ANWB, he will take on the position of CIO at Amsterdam UMC.”

Ferro is a recognized frontrunner in adopting and implementing new technologies. At PostNL, the focus gradually shifted from physical services to leveraging data and algorithms. “Our vision was that data would eventually deliver value,” he explained. This transformation was pivotal in positioning PostNL as a ‘postal tech company,’ emphasizing the importance of in-house data and technology capabilities.

PostNL’s IT strategy has long relied on principles fostering a flexible architecture to adapt to new developments, including AI. The company has consistently stayed ahead of the curve, from fully embracing the cloud in 2013, launching a Data & Insights Competence Center and Advanced Analytics in 2017, to applying GenAI in 2024.

All of this was driven by developments where the volumes of mail and parcels continually shifted places. Data and intelligence were essential to optimize the use of available physical assets. Furthermore, control over the delivery process gradually shifted from the sender to the recipient. The importance of accurate data was further highlighted by the changing relationships with supply chain partners, who were also seeking to capitalize on critical information for their own benefit.

A Successful Journey

PostNL has undergone a successful journey overall. According to Frank Ferro, several aspects remain crucial in this process. Ownership of data initiatives must lie with the business, and organizations should start small and at a manageable scale before industrializing algorithms on a larger scale. Authorized access to high-quality data and embedding robust data governance are also essential.

Ferro also elaborated on the federated structure of internal data capabilities, designed to operate as closely as possible to the business. He highlighted the accelerating impact of a dedicated GenAI task force, all with the aim of creating value as effectively and rapidly as possible.

Aside from data-related content, the closing Q&A raised the question of how leaders and organizations determine which aspects of data and AI to manage in-house and which to delegate to partners. Distinctive processes and activities appear to be the key factors in this decision: ‘Your own intellectual property and what sets you apart from the competition,’ Bubberman and Ferro agreed. ‘Of course, consultants can help clarify this.

Want to know more about becoming an AI-driven enterprise? Read our blog series: How do we become an AI-driven enterprise?