From Java Engineer to Machine Learning Lead: Navigating a Dynamic Career Path

Machine Learning (ML) technologies have become akin to what the web was two decades ago—a transformative force with broad applications. 

Now, you can discover creative uses for ML across various domains, and we’ll undoubtedly see the rise of technology adoption across industries soon. 

We’ve asked Ihar Nestsiarenia, Lead Software Engineer at EPAM Lithuania, to share his career journey from a Java backend engineer to a Lead ML Engineer. 

In the following blog, Ihar shares what he believes are the essential skills ML engineers should possess to excel in their careers and discusses helpful resources and trends that will help ML professionals navigate the future. 

Could you walk us through your journey from a Java backend engineer to becoming a Lead in ML? What were the pivotal moments that shaped your career path?

My journey from a Java backend engineer to a lead ML engineer has had several key moments that shaped my career path. 

It all began during my university years when I joined a research team in mathematical modeling. Soon, my interest shifted toward ML, and coincidentally, this transition occurred during the rise of Deep Learning, which further fueled my passion for ML. 

Later I joined EPAM as a Java Engineer, and my work involved developing services related to search and Natural Language Processing, or NLP, which was close to ML.  

After a few years in this role, an exciting opportunity arose—our customer was seeking someone with both data science and engineering experience to work in a production environment. This was the pivotal moment that brought together my interests in research and engineering.  

Since then, I've encountered opportunities in both technical and managerial paths. I find great satisfaction in being an individual contributor, especially when I have strong managers and clear goals to guide me. I find it really motivating to understand the "whys" behind our solutions and the "how" they benefit the intended audience. 

In my current project, my focus is on transitioning ML development from a startup setting to align with corporate standards. Additionally, I am working on optimizing ML processes within a customer's organization. While creating a single ML model is straightforward, transforming it into a fully-fledged AI company requires integrating tools and establishing clear processes for ML.  

As a Lead ML Engineer, you likely get involved in hiring decisions. What qualities do you look for when selecting new members for your team?

First and foremost, I look for passion in candidates. It's not just about their ability to solve problems; it's about the enthusiasm and commitment they bring to their work. Understanding is another crucial trait I seek. I evaluate the depth and breadth of a candidate's knowledge: do they provide the right answers, or do they exhibit a deeper understanding of navigating complex problems? 

Other vital attributes are professional integrity and open communication. I highly value candidates who are not only vocal about their solutions but also willing to share their concerns with the team. One more quality that I appreciate in candidates is curiosity. I look for individuals who display an inquisitive nature beyond the scope of their immediate project responsibilities. Finally, a commitment to continuous learning is non-negotiable in the rapidly evolving field of ML. Staying updated with the latest research and tools is essential for long-term success.

From a technical perspective, what skills do you believe are essential for aspiring ML engineers to develop to excel in their careers? 

From a technical perspective, qualifications in the ML field are not defined by a specific list but rather by a highly interconnected graph of skills. These skills encompass building strong engineering fundamentals, including proficiency in programming languages like Python, along with a deep understanding of software development principles, version control and code optimization. 

Additionally, expertise in data engineering is essential, involving the ability to work with large datasets and familiarity with tools and frameworks such as Apache Hadoop, Spark, or distributed databases. A solid foundation in data science concepts, covering statistics, data preprocessing, feature engineering, and model evaluation techniques is also a necessary component. 

With the prevalence of cloud-based solutions, competence in using platforms like AWS, Azure, Google Cloud, or others is increasingly important, and familiarity with cloud-based ML services can be a significant asset. Lastly, knowing how to deploy ML models into production environments is critical, and skills in containerization, orchestration (e.g., Docker and Kubernetes) and deploying models as APIs are valuable for success in this multifaceted field.

Continuous learning is essential in such a rapidly evolving field. How do you stay updated with the latest research and incorporate it into your work?

To stay informed about the latest developments in the field, I employ several strategies. I regularly follow data science research to keep up with the latest advancements and breakthroughs. Additionally, it's crucial to keep track of cloud providers like Databricks, AWS Sagemaker, GCP Vertex AI, and Azure Machine Learning Studio, where you can access training materials and earn certifications. I 

Furthermore, I actively share knowledge by participating in tech talks, serving as a speaker, and even creating local communities since joining EPAM. These activities not only facilitate networking but also encourage self-reflection. When preparing talks, I distill my experiences and gather valuable feedback, which helps me enhance my knowledge and develop skills in structuring information, effective presenting, and clear communication – all of which are vital for today's engineers.

Looking ahead, what trends do you foresee in the next few years that aspiring ML engineers should prepare for?

Drawing from my experience, two significant trends stand out: the rise of Large Language Models (LLMs) and the importance of adopting a problem-centric approach. LLMs like GPT-3 represent a transformative leap in the field, with even more innovations expected in the emerging field of LLM Ops. As specialists, it's crucial to be ready to leverage these new ML tools effectively, akin to transitioning from horses to engines for intellectual work. 

Additionally, it's vital to focus on solving problems rather than becoming overly attached to specific technologies, given the dynamic nature of the technology landscape. While predicting the future of technology is challenging, assessing knowledge's lasting value is crucial. Striking a balance between foundational and emerging skills is vital, as enduring knowledge remains valuable, but staying open to new developments is equally essential. 

In summary, ML engineers should prepare to embrace advancements in LLMs, adapt to evolving toolsets and maintain a problem-centric approach, ensuring their success in the evolving landscape of the field.

If you want to join our team, view career opportunities at https://www.epam.com/careers.