Like machine learning engineers, machine learning scientists are in high demand in today’s job market. That’s because organizations are eager to adopt machine learning-powered tools to enhance the value of their data and analytics and add automation to processes.
Demand for machine learning technologies is on the rise, according to market research. Potential applications include customer segmentation and investment prediction in the financial services sector; image analytics, drug discovery and personalized treatment in healthcare; and inventory planning and cross-channel marketing in retail. But machine learning can be used to enhance processes in virtually every industry.
Naturally, there is a need for people who are experts in machine learning and related disciplines, and who understand how to use the technology for practical applications. Machine learning scientists certainly fit that description.
What a machine learning scientist does
Machine learning scientists share many of the same responsibilities as data scientists, including data analysis and model building. Machine learning scientists also work closing with machine learning engineers. A machine learning scientist focuses on researching complex algorithms and building models. Machine learning engineers turn those models into products.
To find out what’s involved in becoming a learning machine, we spoke with Amy Steier, principal machine learning scientist at the developer tools provider, Gretel.ai.
Becoming a machine learning scientist
Steier received a Bachelor of Science degree in computer science from the University of California at Santa Barbara (UCSB). She then went on to earn a PhD in computer science from the University of California, San Diego (UCSD), with an emphasis on artificial intelligence (AI) and machine learning.
A career in technology was not a certainty during college years, however. “I was originally a bit torn between psychology and computer science,” Steier says. “But since I leaned a bit more towards computers, I decided to major in that. I massively enjoyed it and never looked back.”
Math had long held an interest for Steier. “In my early school days, I was good at and very much enjoyed math,” she says. “It felt like a game to me. In high school, I was encouraged by my teachers to join the math club, so eventually I did. All of my friends found this hysterical.”
Steier started to grasp the idea that people must have an inherent tendency to enjoy what they’re good at. “This belief was later a big motivator in my decision to go to grad school,” she says. “I reasoned that if I was going to devote so much of my life to my career I should try and enjoy it as much as possible, and one way to do that was to get very good at something.”
During graduate school, Steier became passionate about data science and specifically about the power and potential of data. “Data science has always been a very fast-moving field, and to remain good in it requires constant learning,” she says. “My passion for the field makes me constantly want to learn and experience more.”
Early education and employment
After graduating from UCSB, Steier’s first job was as a programmer analyst at Computer Sciences Corp. (CSC) in 1986. At the time the company was building a large financial system for the US Navy. “The work was satisfying, but it felt as if I was learning so much more about Navy finance than I did about computers,” she says.
With the goal of refining her expertise, and thus enjoying her work more, Steier went to graduate school in 1990. After exploring different topics, she focused on UCSD’s Artificial Intelligence Group, and was able to work part time at CSC for the first two years .
Following that, Steier took a position as a consultant to Encyclopedia Britannica in 1992, and was able to use the Encyclopedia’s data in her PhD research. “The data they had was stunning in its richness and untapped potential,” she says. “Thus began my enduring, passionate love affair with data that would last my entire career. The power of it, the mystery, the intrigue, the potential has always fascinated me.”
After Steier earned a PhD, she became a director of research and development and then eventually vice president of research and development at La Jolla Research Lab.
On following her passion for data
In 2000, Steier took about a year and a half off for the birth of her son. She eventually started back up part time as a consultant for ContentScan, doing intelligent bibliographic analysis. From there, she took a job in 2003 working part time at Websense. She worked in and eventually ran the CTO office, exploring new technology and product directions.
“At that point in my career I was faced with a big decision,” Steier says. “Do I stay on a path of management or redirect myself to focus more on hands-on work? I loved being able to set a vision for a group and to help team members flourish in their careers. But I was passionate about hands-on work. I followed my passion and have never regretted it. Even today, when asked for advice from someone on what career path to follow, I still advise to follow your passion.”
Steier took on a role at Websense as lead researcher on a classification system for the web. “We primarily used large support vector machines to classify content into more than 80 topics and a dozen different languages,” she says. “That system is still in use today.”
When cyber security became a hot topic and Websense—eventually bought by Raytheon and now called Forcepoint—transformed into a security company, Steier took a role in the cyber security group. “I became involved in a plethora of innovative projects focused on both web and data security,” she says. “I worked on automatic classification of malware, detection of outbound malware communication, automated detection of malicious web sites, visualization of the threat landscape and other innovative projects.”
In 2019, Steier had lunch with a former colleague who was on his second successful startup venture. “When he explained the mission and vision of Gretel.ai, I was instantly hooked,” she says. “The mission was to remove the privacy barrier to sharing data for everyone. Easy access to data had been the thorn in my side for as long as I could remember.”
Joining Gretel.ai “was like coming home,” Steier says. “My career has always been driven by my passion for data, and now I am able to focus on helping everyone harness its power and potential.”
A day in the life of a machine learning scientist
“I like to start my workday by looking over my reading queue and seeing what’s either interesting or relevant to read that morning,” Steier says. “Then I usually have a couple meetings each day—either on company or research team related topics. I try to keep my meetings grouped together so I can have focused time on whatever research project I’m currently on.”
Sometimes that work involves more reading to explore what has been done so far or searching for technology innovations that might inspire some new angle to a project. Steier spends a good deal of the day building various proofs of concept, each connected to a vision in the company’s product roadmap.
“We’re currently hiring, so once a week I’ll have a phone screen or interview,” Steier says. “We write a lot of blogs, give interviews, do podcasts and talks, so I might spend some time on one of those items. Maybe once a month I’ll get involved with a specific company’s use case and help [plan] a solution. We chat a ton on Slack on both work-related topics and random interesting or amusing topics.”
Career defining moments
We asked Steier about her most memorable career moments. “What really stands out was the aha moment at Encyclopedia Britannica, when I realized my profound love and fascination for data,” Steier says. “I can remember the exact moment I was explaining it to a colleague at a conference. Saying it out loud made it really sink in. I’ve carried that passion with me throughout my career.”
More recently, “joining Gretel has caused me to become re-energized about my passion for data and what it is enabling spaces in the machine learning and AI,” Steier says. “When I first started working within the world of data, a lot of what companies were doing was hindered by the inability to access or share data due to privacy concerns. But I’ve been watching this change in real time thanks to synthetic data. Tools, like what we are building at Gretel, remove barriers and allow data to become ever more democratized. I see this as enabling tech communities across the world to utilize more datasets and harness the power they provide.”
Getting a PhD also opened a lot of doors, Steier says. “After that, continued learning became just a natural and necessary part of my career,” she says. “This has always meant lots of reading, communication with colleagues and being open to trying out new ideas.”
Inspirations and advice for others
Her parents were her biggest inspiration, Steier says. “During most of my life, my father was a professor of electrical engineering at USC [University of Southern California], and my mother owned several clothing stores. It was always clear they enjoyed their work. Going to college was never a question, just a natural part of growing up. Having the courage to push forward and go to grad school was solidly based on my parents’ unwavering faith that I could accomplish that.”
“No life is without hardship, but I believe my passion in my work has helped me to be resilient,” she says. “Through every loss of a loved one, my work provided a refuge that helped me regain my footing.”
For others seeking a path similar to her own, Steier’s advice is simple: “Get educated, follow your heart, and continuous learning,” she says.