Jeff Schwartzentruber of eSentire On Five Things You Need To Create A Highly Successful Career In The AI Industry
Make AI accessible to everyone. We want to have as much diversity within the AI industry as possible, and we want it to be accessible to everyone. There are the traditional STEM paths, but really you can come from any academic or cultural background — you just have to be willing to learn and ask questions.
Artificial Intelligence is now the leading edge of technology, driving unprecedented advancements across sectors. From healthcare to finance, education to environment, the AI industry is witnessing a skyrocketing demand for professionals. However, the path to creating a successful career in AI is multifaceted and constantly evolving. What does it take and what does one need in order to create a highly successful career in AI?
In this interview series, we are talking to successful AI professionals, AI founders, AI CEOs, educators in the field, AI researchers, HR managers in tech companies, and anyone who holds authority in the realm of Artificial Intelligence to inspire and guide those who are eager to embark on this exciting career path.
As part of this series, we had the pleasure of interviewing Jeff Shwartzentruber.
Over his 10-year career, Jeff Shwartzentruber has been involved in applying machine learning models for threat detection and security analytics for several large financial institutions, public sector organizations (federal) and subject matter experts, and his current work as Senior Machine Learning Scientist at eSentire involves developing customized LLM models that intelligently enhance a Security Operations Center analyst’s capabilities when performing security investigations for the Managed Detection and Response provider. In addition to private sector work, he is an Adjunct Faculty at Dalhousie University in the Department of Computer Science, a Special Graduate Faculty member with the School of Computer Science at the University of Guelph and a Research Fellow at the Rogers Cybersecure Catalysts. He holds a PhD in Mechanical Engineering from Toronto Metropolitan University, focusing on analytical process modeling.
Thank you so much for joining us in this interview series! Before we dive in, our readers would like to learn a bit about your origin story. Can you share with us a bit about your childhood and how you grew up?
I was raised on a farm near Hanover, Ontario in Canada, and I feel that I was never really meant to go into technology. Growing up on the farm, I learned that I love to build and work on things, so I pursued engineering. I’ve always been very hands-on, fixing things, figuring things out, etc., and I found myself wanting to apply these skills in a technical field.
Can you share with us the ‘backstory” of how you decided to pursue a career path in AI?
As I finished my undergraduate degree, I wrote my thesis on genetic algorithms for winglet designs. While working on my thesis, I realized that its core engineering is a very iterative process if you don’t use machine learning. For instance, you come up with a design, you test that design, and then, based on your expertise or analysis, tweak the design, Then you redo that process until you feel like you’ve reached a level of engineering you’re confident in. But when working on this genetic algorithm winglet design for an aircraft, I developed a program that did this iterative process faster and, in some cases, more intelligently. The result was a winglet design far superior to the one I would have generated manually in the same amount of time, and I had the data to prove it. The program would go through thousands of designs and then produce an optimal solution given the constraints. I was able to produce the perfect win-win in a very short amount of time. Then, I became hooked on the idea of Artificial Intelligence (AI) and Machine Learning (ML). I continued to apply and advance these AI concepts through my graduate research. While working on my Ph. D., I started working with the startup AI community, further accelerating my progression.
Can you tell our readers about the most interesting projects you are working on now?
I’m working with a lot of generative AI modeling right now. Since our data is very customer-centric, eSentire has had to embrace building, hosting and testing its own LLM and NLP models. Currently, my most interesting project is building an internal LLM that can read documents, assist analysts and help upscale our security capabilities inside the security operations center (SOC) with these self-hosted LLMs. It takes an enormous amount of collaboration between the SOC, data science and engineering teams, and the number of potential applications in the pipeline is very exciting.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful for who helped get you to where you are? Can you share a story about that?
I previously worked with a CEO who was very supportive of innovation, and in doing so, he was very supportive of my learning and growing within the AI field. If I wanted to go to a conference or start a research collaboration with a particular university, he would encourage those initiatives. It was amazing because I was able to really ramp up my research and my connections within the field, which really accelerated my AI career.
As with any career path, the AI industry comes with its own set of challenges. Could you elaborate on some of the significant challenges you faced in your AI career and how you managed to overcome them?
Currently, one of the most significant challenges within the AI industry is security and bias. Because most data is so human-centered and personal (such as conversational text), it makes data analysis very difficult because you have a moving center of truth. This also applies to security, where one person may see something as a particular threat while someone else may put a bigger emphasis on a different risk, so security really depends on the organization and their perspective of security. And this effect of bias creates significant friction with the traditional methods of security analysis. ML is all about probabilities and making predictions based on the most probable outcome, whereas cybersecurity has always been rooted in deep deterministic security research. For example, when developing a detection for malware, the traditional method can answer questions like: this is what the malware does, this is what you need to look for, and this is how you need to protect against it. Whereas a large majority of ML security analytics tries to find when things are anomalous but with much less causality. It becomes a difficult line to walk — it’s like speaking two different languages, the security language and the AI language, and the challenge is getting them to work together.
Ok, let’s now move to the main part of our interview about AI. What are the 3 things that most excite you about the AI industry now? Why?
- Generative AI — I’m excited by all of the different facets of generative AI and just how big it is going to massively change some industries, and we are just getting started!
- The potential growth of AI — We are really just in the infancy stage of some of these language models (like ChatGPT), so I’m excited to see how much better these models will be in the future. In five years, my bet is you’ll have a language model running locally on your phone, you won’t need an internet connection to have all this information available to you. So the implications of all of that are really exciting to me.
- Quantitative vs. qualitative research — I’m a numbers person, but there is a whole side of research that is qualitatively focused, so I’m really excited to see how both of these two worlds side will work off each other and how qualitative will come in to play a significant role in AI research.
What are the 3 things that concern you about the AI industry? Why? What should be done to address and alleviate those concerns?
- The perception of AI — AI’s perception really isn’t great, and I actually wish the term AI didn’t exist. I like the idea of data science because it puts people in a different mindset. When people hear the term AI, it gets blown out by marketing jargon and a lot of misconceptions and it moves further away from the true science behind it.
- Accessibility of generative AI models — Currently, my biggest concern is around the accessibility of generative AI models and the implications of unintended and malicious use cases. For example, the ability to create very high fidelity bots that can do widespread social engineering is becoming increasingly easy to implement.
- Security — Security will always be a concern. Cybercrime is always evolving, and really good generative AI is creating some new cyber risks. For example, AI has lowered the barrier to entry for novice hackers to become effective threat actors. A simple example of this is the ability to write convincing phishing/Business Email Compromise emails. If the victim engages, GenAI can craft the perfect response in the target’s native language. So we have to attack AI security head on.
For a young person who would like to eventually make a career in AI, which skills and subjects do they need to learn?
My biggest piece of advice is you have to keep building things, even if they fail. Just keep building because that’s how you learn. It’s also good to immerse yourself in the field academically — you want to be able to understand where the field has been and where it is going. It’s an intense field, so constantly striving to advance your knowledge of mathematics (such as calculus, statistics, to discrete algebra) will serve you well. Lastly, you really want to be humble — it’s almost impossible not to be considering the breadth of domain knowledge — so ask questions and be passionate about learning.
As you know, there are not that many women in the AI industry. Can you advise what is needed to engage more women in the AI industry?
There definitely needs to be more women in the AI industry. One way to engage further within AI is to combat some misconceptions. There is a current cultural perception that AI is all technical and that you have to be a very technical person to be able to engage in the field. But there are actually a lot of parts of AI that don’t require that technical expertise, so we need to change that perception to make AI learning more accessible to everyone. We need to start breaking down that perception and hyping up its accessibility, and start at a very young age.
Ethical AI development is a pressing concern in the industry. How do you approach the ethical implications of AI, and what steps do you believe individuals and organizations should take to ensure responsible and fair AI practices?
This is an industry concern right now, and unfortunately, the regulatory bodies aren’t going to be able to keep up with the current rate of technology advancements. And it’s unfortunate because that’s where you’d get the biggest support in ensuring its ethical use. AI is progressing so quickly that the internet is a different space than a few years ago. I don’t think at this point we can count on the government to regulate AI (successfully), so the onus really falls to organizations, researchers and developers. We can accomplish this by improving oversight and transparency. Take, for example, a research board at a university. They typically are used for something like a human or animal trials where you submit a research proposal, and a committee of experts reviews it for ethical or moral concerns. This type of approach provides safeguards to make sure you’re continuing research in an ethical way. It may seem like AI doesn’t need this because we’re dealing with computers and numbers, but nothing could be further from the truth. We need to fill the current gap, which means we need some sort of oversight committee.
Ok, here is the main question of our interview. Can you please share the “Five Things You Need To Create A Highly Successful Career In The AI Industry”? If you can, please share a story or an example for each.
1 . Make AI accessible to everyone. We want to have as much diversity within the AI industry as possible, and we want it to be accessible to everyone. There are the traditional STEM paths, but really you can come from any academic or cultural background — you just have to be willing to learn and ask questions.
2 . Strong mentorship and communication. The AI field is evolving at such a fast rate, so when people show interest in the field, we want to be enthusiastic towards them and build them up (that’s also how we address and solve the diversity issue within the industry). I’ve been fortunate in my career to have very supportive and enthusiastic mentors and it has really helped me achieve success.
3 . Never stop learning. As we’ve discussed, the field of AI is so broad, so you can never stop learning and advancing your skills — reading, staying on top of trends, joining communities etc. For me, this is where my network has been really helpful. For example, I’m associated with a few different labs in Canada that focus on the intersection between AI and cybersecurity. I attend a lot of conferences around those topics and am lucky enough to be involved in a lot of industry conversations, so I’m always learning. Being in a culture of innovation and research helps me stay up-to-date.
4 . Be willing to fail. AI models are constantly changing, and what was new and innovative two years ago may not be great today. So you have to be willing to try and test things and then pivot if they don’t work.
5 . Prioritize security. Like AI, cybersecurity is a dynamic industry, constantly changing. And as the saying goes, you’re only as secure as your weakest link. With AI and cybersecurity so intertwined, you need to be up-to-date on the latest cybersecurity trends and be able to work within AI with a security mindset. AI should be rooted in security.
Continuous learning and upskilling are vital in a dynamic field like AI. How do you approach ongoing education and stay up-to-date with the latest advancements in the AI industry? What advice do you have for those looking to grow their careers in AI?
The key to growing your career is growing your network. As I previously mentioned, I participate in several industry associations that help me grow my network (and ultimately my career), attend conferences, and read the latest research/papers, all of which ultimately help give me a competitive edge within the field.
What is your favorite “Life Lesson Quote”? Can you share a story of how that had relevance to your own life?
In order to predict the future, you have to create it.
This goes back to one of the best life lessons within my AI career, which is you have to build things. And keep building. This is the same when it comes to AI and ML…if you want ML to predict the future, you have to make it. In my life, if I’m working towards something, I have to get my hands on it. If I have a question, I get my hands going in that direction — start putting thoughts to code, putting pen to paper, putting nail to hammer. And then from there, seeing where it goes. That is how you learn. And if you tear it back down and have to start over, that’s ok because look at all the learning you got during that time.
You are a person of great influence. If you could start a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. 🙂
I like helping people learn and making learning accessible, so my movement would be around more education. I’ve seen firsthand the power of education. My dad is a ninth-generation farmer and I’m the first in my family to receive a PhD and complete that level of education. Reflecting on my life, I feel like I wasn’t supposed to take my career in this direction, but I’m so glad I did.
How can our readers further follow your work online?
This was very inspiring. Thank you so much for joining us!
Jeff Schwartzentruber of eSentire On Five Things You Need To Create A Highly Successful Career In… was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.