AI/ML: Having a great depth and breadth of knowledge regarding artificial intelligence and machine learning algorithms, and staying up to date, would be a challenging full time job in itself. Most of us in the field know that, while it is usually our favorite part of the day, it is only one of the many skills we use on a daily basis. This is what I do, what I read, and what I think about in my free time. It's the reason I work extra hours; because I can't stop thinking about it and every day feels like the whole AI world has changed. 

Software: While us ML folks don't usually call ourselves software engineers, we do spend a ton of time designing and writing software. I myself thoroughly enjoy software engineering and, like many others, have found a rigorous software development process can bring much needed order to the non-deterministic world of machine learning.

Leadership: I am very fortunate to have already held many leadership roles in my young career. I find that my impact is magnified and my natural social strengths are showcased when I'm leading teams to develop AI/ML and software solutions for the world's toughest problems. I am even more fortunate to have had so many wonderful people on my teams, and being able to harness the potential of each individual on my team is a responsibility I take very seriously. A final thought on leadership: because excelling in the fast paced AI world puts you in so many circles with so many people, good leadership from any position can be extremely valuable!

Business Development: In most of my roles, I am not simply handed some funding, a well constrained problem, and told to solve it. It is often the case that someone has a problem and doesn't know exactly what solution they need. This is where business development skills are required. Without having someone who can pitch a variety of solutions, their pros and cons, and what they would cost, the customer is usually left dissatisfied. I find this part fascinating, as you really have to put yourself in their shoes to help them reach their goals. 

Application Expertise: This is an interesting phenomenon. One would think that being a machine learning engineer means you don't have to learn how to do phased array radar math or understand the decision making science behind strategic planning. But if you work with people doing those things, you would be wrong. Despite being the machine learning expert on the team, it is incredibly important to get at least a basic understanding of the application space you are working. You will often serve as a translator, re-framing the problem statement to identify what actually needs to be done, often rethinking the entire process that existed before machine learning was involved. Neglect this piece at your own risk!