Garfield Refining

Kyle Abrahams

AI Leadership Visionary

Kyle Abrahams is Senior Director of Technology at Garfield Refining, where he leads digital transformation, software modernization and technology strategy initiatives. With extensive experience in technology leadership, process improvement and business-focused innovation, he focuses on building solutions that enhance operational efficiency and support long-term growth.

As I moved into leadership, I was determined to assist everyone by fulfilling every request that came my way. Gradually, it became apparent that some of those requests were making an employee's life easier but had no added value to the organization or its clients. Essentially, those requests were draining my department resources, preventing more meaningful projects from moving forward in a timely manner. That forced me to take a step back and re-evaluate my entire approach to leadership.

My leadership approach is to work with stakeholders to ensure projects are business-focused and that my team is being well-utilized. I constantly ask myself, “Does this ultimately serve the business?”.  When my team is supported and working well, they make our employees more effective, and that allows the business to grow and deliver for our clients.  That chain, from team to employee to business to client, is the framework I return to whenever the noise gets loud. It helps my team stay focused on work that actually moves the needle rather than busy work that fills up the day.

Establishing the Right AI Foundation

The foundation has to be solid, built upon efficient processes and clean data. AI is an accelerator at its core, which will magnify what is already there. If your processes are inefficient or your data is producing junk, AI is going to make that problem bigger, faster.

Beyond the foundation, governance is the piece most organizations underestimate. That conversation has two sides. The first is data exposure: understanding what information or intellectual property you are potentially sharing with your vendor of choice. The second is harder to see coming: AI can perform unauthorized actions in service of what it believes to be the end goal. We have already seen real-world examples of AI taking irreversible actions against a user's explicit wishes. Governance needs to clearly outline what data you share and what level of autonomy you allow.

Modernizing With Purpose and Business Value

Value drives my decisions. What is going to reduce the most friction? What is going to move the bottom line based on realistic projections? If there is low-hanging fruit to pick up along the way as we modernize, we will take it. But modernization for modernization's sake is rarely the goal.

  Transformation is not about modernization alone. It is about creating business value, empowering teams and focusing technology resources on what truly drives growth.  

When there is organizational pressure to modernize something because it looks outdated, I sit down with the stakeholder and ask them how the change is actually going to bring value. How does the end picture differ from the current one in a meaningful way? That conversation either builds the case or it doesn't. Either outcome is useful.

Driving Adoption through Effective Change Management

Two things stand out. First, employees need to be ready for the change. If you push transformation before the people are ready, you can create a backlash that sets adoption back years. The technology can be right, but the timing can still be wrong.

Second, there is no such thing as over-communicating. Communicate the why behind the change. Communicate what is coming and when so nobody is caught off guard. Do both frequently and often. The organizations that struggle with transformation are rarely struggling because of the technology. They are struggling because people did not understand what was happening or why it mattered to them.

Preparing for the Next AI Era

AI is already nearly everywhere, and the real challenge ahead is prioritization. AI tokens are currently subsidized from a software perspective, and the infrastructure needed to support AI demand is lagging behind. Those two factors are going to converge and create a significant supply issue. Organizations will be competing for compute time, and that compute will come at a premium.

The leaders who are positioned well when that happens will be the ones who built governance and prioritization frameworks before the crunch hits. Figuring out when to say no to AI now, while the cost of getting it wrong is still low, is exactly the work that will pay off when the stakes are higher. AI is not a one-size-fits-all situation. Understanding where it actually matters for your business, and where it does not, is a competitive advantage that will separate the prepared from the scrambling and avoid unnecessary costs to the business.