Lt. Col. Pedro Ortiz received his commission in May 2004 after earning his Bachelor of Science Degree in Economics from the Wharton School of Business at the University of Pennsylvania. In January 2005, he completed basic officer training at The Basic School, Quantico, VA, with follow-on training as a Communications Officer at Communications School, Quantico, VA. Upon completion of MOS school, Ortiz was assigned to 7th Marine Regiment where he deployed to Iraq as the Regimental Communications Maintenance Officer. In 2007, Ortiz deployed to Iraq for a second tour as the 1st Battalion, 7th Marines Battalion S-6 Officer. In June 2008, Ortiz received orders to NPS as a master's student in the Computer Science Department. He graduated early with an Outstanding Thesis distinction, and received follow-on orders to the United States Naval Academy as an instructor in the Computer Science Department.
Ortiz other assignments and deployments include Combined Joint Task Force-Horn of Africa, 2d Marine Expeditionary Brigade, and the Marine Corps Warfighting Laboratory (MCWL). He also attended Marine Corps Command and Staff College. His next assignment is with the DoD Chief Digital and Artificial Intelligence Office (OSD CDAO).
The greatest impact on my time here was the people. I came to NPS during the first summer of COVID. There was a significant amount of uncertainty about many things as the whole world adjusted to the reality of the situation. I found a group of individuals that were willing to be invested in my success regardless of a global pandemic. I had a great advisor in Prof. Marko Orescanin as well as three stellar atmospheric scientists, Prof. Scott Powell and Dr. Eleanor Casas from NPS and Dr. Veljko Petkovic from the University of Maryland. There were also supportive Marine leaders, such as Colonels Jason Perry and Randy Pugh and Lieutenant Colonels Emmy Hill and Amy Roznowski, that made sure I had the support I needed to grow not just in the halls of academia but as a Lieutenant Colonel supporting fleet units as part of the Fleet Engagement Program.
During my dissertation defense, I had the tough job of condensing over two years of complex, technical work into a 40-minute presentation that could be understood by a wide audience. This was challenging to say the least, but it was also very applicable to what it means to be a technical leader. Part of leading is influencing both your superiors and subordinates. The ability to take complex ideas and communicate them succinctly is a must-have for anyone who want to lead technical efforts, particularly when your boss or people that work for you may not be very familiar with the technology that is being developed. This exactly why the Marine Corps created a technical PhD program, and why I was sent to NPS.
One topic that has some growing interest across the DOD is model monitoring. My work in uncertainty and deep learning highlights one way that AI models could be monitored. The whole point of monitoring a model is to makes sure it is behaving as expected or if something needs to be updated or changed. Uncertainty provided by probabilistic models makes the model output more interpretable and allows for better decision-making regarding the performance of a deployed model. These methods are rarely used by DOD systems, if at all. It is my hope that we will see these types of models being leveraged more frequently as the DOD develops AI and AI-enabled systems.
One of the great opportunities that NPS afforded me was that my research was interdisciplinary from the start. A team of just computer scientists or just atmospheric scientists could not have made the contributions to science that our team, comprised of both, made. As a computer scientist, I need the use cases that experts in other fields deem important. Otherwise, I would be working on a problem that computer scientists think is interesting but may have little or no value in the real world. As a computer scientist, I can bring to bear cutting edge technology that is inaccessible to researchers in other disciplines because the hard parts are really hard and require a high level of technical expertise. I believe that the interdisciplinary work that is possible at NPS could advance the development and implementation of AI technologies in defense by working on data that that is uniquely relevant to the DOD and that is potentially only available in classified environments.
