For organizations to fully realize the promise of AI, they must trust the data it uses. That’s why knowledge management, the practice of identifying, capturing, organizing and optimizing critical institutional data, is so important. Newly appointed Kyndryl Distinguished Engineer Pedro Soares leads the firm’s efforts to transform unstructured data and undocumented expertise into actionable information. Drawn from documents, repositories, processes and people, this knowledge can become the source of truth for an organization’s AI-enabled IT estate.
Ultimately, Pedro’s work centers on trust. Here, he explains how data-driven transformation provides the foundation for scalable innovation and business growth.
Can you describe the importance of Enterprise Knowledge Management (EKM) to the uninitiated?
Pedro Soares: EKM is not a new concept. Organizations have always needed to locate, understand and manage their intellectual property and business processes. What has changed in the era of AI is the greater need to make that data accessible, quantifiable, consistent and governable. AI has illuminated the gaps and shortcomings in traditional EKM and has shown us that we need to consolidate disparate pockets of data into taxonomic structures with clearly defined relationships among the information — an ontology. We know that everything we need is in the garage. Now it’s time to clean it up, put our tools in the right boxes, label the boxes clearly, and maintain the relationships between all those pieces.
Once we have correctly structured and governed our data, we can use AI to generate recommendations, create specialized knowledge packages, establish and maintain information lifecycles, and produce outputs we can trust. We know that AI is only as useful and safe as the data it draws from. EKM is the science of making that data useful.
How is Kyndryl helping enterprises infuse AI into their operations, and what are some specific reasons for organizations to undertake these transformations?
Soares: We’ve established that organizations need to prepare their data and infrastructure for AI infusion. But they also need to modernize and, in some cases, redesign their business processes before automating them. Otherwise, they risk automating flawed processes that AI will amplify at speed. That’s why Kyndryl begins with the data. We employ our platforms and capabilities — Kyndryl Bridge, Responsible AI at Kyndryl, Kyndryl Agentic AI Framework — to help customers get started on their AI journeys. But technology in a vacuum is meaningless. Our customer engagements begin with consultation and listening. We work to understand each organization’s operational challenges and ultimate goals, then apply that knowledge and perspective to the technology solutions we design.
Kyndryl’s modernization demonstrates that transformation is possible and that we have the expertise to deliver it. We also continue to develop unique capabilities based on our foundational experience, decades of work across every generation of enterprise IT and the cross-industry data we analyze each day.
AI is only as useful and safe as the data it draws from. Enterprise Knowledge Management is the science of making that data useful.
Distinguished Engineer, Vice President, Knowledge Management
What passions drew you to this line of work?
Soares:One might say that I’m a bit obsessed with structure, which led me to study engineering. Although I wouldn’t say that coding is my main interest, I’ve always been fascinated by compilers and the way people can communicate with machines. I began experimenting with neural networks in the 1990s and have been fascinated with building digital structures that enable people to access and reuse knowledge ever since.
I also enjoy developing talent and building teams. I’ve had the privilege of leading architect communities at the country, regional and global levels. Helping people grow energizes me.
Those experiences also prompted me to consider structural challenges. What happens when organizations rely too heavily on human memory and scattered pockets of data? They end up with pieces of a puzzle but never the complete picture. Businesses can’t scale when their data remains disconnected, or they expend human energy on mundane, repeatable tasks. The goal is to have AI handle the process-intensive work so people can focus on matters requiring discernment and creativity. The key to both outcomes is assembling a useful data ontology that people can trust.
What advice would you give to your younger self?
Soares:Listen more! Active listening means absorbing what others are sharing and teaching instead of jumping to conclusions or simply waiting for your chance to talk. It’s one of the most valuable skills you can have in your professional or personal life. I also would have advised myself to be resilient, be patient and seek feedback. Always keep learning.
That doesn’t mean accepting everything you hear without evaluating it. As a mentor advised me, it’s important to listen and seek feedback, but it’s just as important to evaluate what you hear critically. That’s where resilience and patience come into play. As you gain experience, you become more skilled at listening, synthesizing information, and working with others to create value.
What activities do you enjoy in your spare time?
Soares:My family is the most important thing in my life. I have two children, who aren’t little anymore, who keep me engaged, challenged and always learning. We live on the western Portuguese coast and enjoy traveling, swimming and water sports.
I also enjoy martial arts, scuba diving, making and listening to music and all sorts of reading, including fiction, nonfiction and technical material. But my greatest passion and my therapy is motorcycling. I take short weekend trips when I can and complete at least one long tour with friends every year. Along with offering freedom, scenery and new experiences, responsible motorcycling requires preparation, discipline and concentration. Riding forces me to focus completely on the present and disconnect from daily pressures.
