Becoming an AI Engineer: 7 Questions with an IBMer

Before becoming an AI Engineer, Texas A&M alum Isaac Ke held the titles of statistician, data scientist, and machine learning engineer. Today, he is a subject matter expert in building generative AI systems with foundation models. Join us in the following article as we discuss both the role of an AI engineer and AI in the modern workplace.

AI, everywhere, all at once

From 2022 to 2023, the percentage of Americans more concerned than excited about AI in daily life jumped from 38 to 52 percent, according to the Pew Research Center [1]. This shift in public perception can be partially attributed to the increased accessibility and implementation of AI tools. Many websites now offer their own versions of AI, including Adobe’s Firefly, X’s Grok, Snapchat’s My AI, and Spotify’s AI DJ. 

Concern over AI’s growing omnipresence is unsurprising, but it isn’t unwarranted. Employers across industries not traditionally associated with tech, such as retail, have begun prioritizing AI literacy [2]. According to a 2024 work trend report, 66 percent of leaders say they won’t hire someone without AI skills [3]. Apprehension is valid, but here are two reasons why you shouldn’t let it dampen your excitement:

• Due to AI’s widespread nature, many of us share the same learning curve. A great deal of the AI applications you encounter in daily life are experimental, meaning employers are still learning how to use it, too. 

• You don’t need to become an expert in every emerging AI model. With so many AI tools and applications, how can you know where to start and which ones will be meaningful to your work? Here’s Ke’s take on how to approach AI upskilling: 

"Since data science and AI engineering can be applied to any industry, it’s important to learn more about the specific domain you’re seeking to work in. ... AI engineers and data scientists across different companies and markets will have varied perspectives on the field’s outlook, definitions, and focuses — all of which are valid and applicable for each area."

Read more: How to Learn Artificial Intelligence: A Beginner’s Guide

A day in the life of an AI engineer

If you’re not familiar with AI engineering or could use some additional context, here’s a quick breakdown: It’s the use of artificial intelligence and machine learning (ML) to develop systems that solve problems and increase efficiency. Depending on the industry and company they work for, AI engineers might:

• Build AI models and program applications.

• Conduct statistical analyses and present results to help guide decision-making in an organization.

• Implement AI infrastructure to streamline or automate development and production processes.

Read more: What is an AI Engineer? (And How to Become One)

Which of your job tasks do you find to be the most engaging?

I really enjoy the open-ended nature of AI engineering tasks. The generative AI landscape is evolving at a blazing fast pace, and there are thousands of ways you can combine different models, prompts, frameworks, techniques, orchestration layers, and UI [user interface] elements to build an AI solution that delivers business value for clients. Every day I get to experiment with new approaches and see what the technology is capable of.

One aspect of my job I’ve grown to enjoy more is the client-facing interactions. As part of IBM Client Engineering, I work directly with clients to co-create pilot solutions tailored to their business needs. Getting to demo the solutions to stakeholders and see their faces light up is one of the many rewarding things about my job.

How do you continue learning and growing in your field?

I like to use the T-approach for learning. This means having a basic understanding of a wide breadth of knowledge, and then strategically choosing a few key areas to have a depth of skills in. Online educators like IBM provide numerous courses and badges to be able to gain that breadth of knowledge. For the depth of skill, I like to browse research papers and articles and get hands-on with code repos [Editor’s note: Code repos or repositories are data structures where software development assets like code and documentation are stored]. I will find Github repos, Hugging Face models, and even new Python libraries to try out. I also enjoy chatting with my peers about what they’re working on in their projects.

What three key things have made this stage of your career possible?

1. Be a challenge taker. One of the first breakthroughs I had in my career was when I had an opportunity to lead a project and present a demo of it to a large audience. With lots of preparation and support from my team, I gradually gained confidence, and the event ended up being a huge success! Taking that first step is the hard part, but having a growth mindset to learn along the way is what propels you to the next level. The reality is that many things are out of our control, so all you can do is be excellent in what you’ve been called to do and be ready to jump in whenever opportunities arise to take on more responsibility.

2. Be a go-getter. Bias yourself to action. Don’t wait for people to tell you what to do to get started. You can always take initiative and be proactive in bringing ideas to the table or taking that first step to learn. And if you don’t know where to start, that’s OK! Ask lots of questions because there’s a high probability that someone else also has those questions.

3. Be a team player. I’ve had the blessing of working on some amazing teams in my career. What I’ve enjoyed is that each person puts the needs of the broader team and, ultimately, our clients first. The trust you build with this mindset helps the team work more efficiently and accomplish what was set out to do. 

Inside an AI engineer’s toolbox

Ke works in IBM’s Financial Services Market, in collaboration with Truist Financial’s Innovator in Residence program. This program creates space for field experts from start-ups and tech leaders to share insights and explore applications for emerging technologies in finance. Examples of use cases for AI in the financial sector include business process automation and new payment technologies. 

What tools have you had to master to succeed in your career?

Python is definitely the coding language I use the most. Whether it’s working in Jupyter notebooks or writing lightweight applications, I use Python to build AI solutions using both IBM watsonx SDK’s [software development kits] as well as open-source libraries such as LangChain and Hugging Face.

I also work in the Cloud a lot, leveraging Saas applications for generative AI components and Redhat Openshift to deploy full-stack AI applications.

Read more: 9 Python Libraries for Machine Learning

Which skills are you surprised that you utilize most often in your role?

I do a good amount of writing documentation for code and diagraming repeatable assets. With the GenAI field changing every day, it’s helpful to enable others to leverage what you’ve built and collaborate with fellow engineers.

Also, being in a client-facing AI engineering role requires me to hone my communication skills. I often have to translate technical jargon into digestible information for non-technical folks, explain points of view and best practices, and facilitate conversations to manage expectations between business ambition and technical feasibility.

Most of the day-to-day skills I use now, whether technical or soft, I learned on the job through trial and error, watching mentors, and being a sponge for knowledge.

Advice for aspirers

The path to success looks a little different for everyone, especially in today’s fast-paced job market. The next couple of questions and answers from Ke shed light on what you can expect in the early stages of a tech career, whether you’re interested in becoming an AI engineer or not. 

What type of education was most helpful as you were beginning your career?

My bachelor’s and master’s degrees in statistics helped provide me with a rigorous foundation in mathematics and computer science necessary for working in the AI field. Learning the technical inner workings of machine learning models gave me the right paradigm and lens to effectively solve business challenges with AI/ML.

Apart from academic training, I learned a great deal by getting hands on with the technology and experimenting by participating in datathons, data science competitions, internships, and badging/certification programs offered by IBM.

Knowing what you know now, what advice would you give to your younger self as you were seeking your first role?

I would give two pieces of advice: One, don’t feel like you need to know everything to start your first role. Most of the day-to-day skills I use now, whether technical or soft, I learned on the job through trial and error, watching mentors, and being a sponge for knowledge.

My second piece of advice would be to keep up to date with the latest trends and tools in the industry. There are new generative AI models, products, start-ups, and techniques being published every single day, and it’s important to know how to talk about the broader landscape when working with stakeholders.

Learn more from Isaac Ke & the IBMers

Can you guess what IBM considers to be their greatest asset? It’s the forward thinkers like Isaac Ke who make up IBM, or, in other words, IBMers. If you already have some relevant experience and you’re interested in learning AI engineering, consider enrolling in IBM’s AI Engineering Professional Certificate program. With the help of expert IBMers, you’ll master concepts like computer vision and deep learning using industry-standard tools like SciPy and Tensorflow. 

Curious minds who are new to the field may instead start with an introductory-level course series like IBM’s Data Science Professional Certificate program. Before moving on to advanced AI skills, you’ll master foundational concepts like predictive modeling and data visualization.