Discovering Data Science Career Paths
Data science is one of the fastest growing STEM careers with high demand for skilled personnel to extract useful information from structured (e.g. databases, spreadsheets) and unstructured data (e.g. text, images, video, audio). Perhaps less well known is that job opportunities are not limited to big tech (e.g. FAANG); rather, companies across sectors as well as the public sector collect data and are building teams to help them derive insights.
Thus, in charting a career path, it is advisable to look beyond tech companies to the myriad opportunities in other sectors where data science is being increasingly integrated into operations. In fact, if job seekers focus narrowly on tech, they run the risk of missing a gamut of job openings for which their skill sets could apply.
Artificial Intelligence and Advanced Analytics at Spartan Controls
Modern industrial processes are utilizing an increasing number of sensors, sophisticated process control systems, and a myriad of other industrial data systems in the safe and efficient operation of these facilities.
These processes generate large volumes of data which creates the opportunity to leverage modern AI/ML technologies to unlock the true potential of existing operational data enabling our customers to increase efficiency, safety and environmental compliance.
Conceptualizing, developing, and commercializing AI/ML solutions will generate significant business value for process industries by leveraging AltaML’s AI/ML development capability with Spartan’s domain knowledge and expertise around industrial processes, automation, optimization, equipment and process reliability.
The largest group of data science and engineering professionals in Canada have joined together to focus on the application of artificial intelligence in the process industries. AltaML and Spartan Controls are excited to work closely in the application of these modern technologies providing data analytic solutions across Canada and the rest of the world.
Applied AI Lab Associate Perspectives
AltaML connected with Associates partnered with Spartan Controls to get their impressions of applied data science career paths in the industrial automation sector, on the importance of domain expertise, and to see if they have any advice to share, which they do, aplenty.
Marzieh Kooshkbaghi, AltaML Associate ML Developer, holds a PhD in Electrical Engineering in Control Systems, and was familiar with data science in the financial field for uses such as risk modeling, fraud detection, customer segmentation, real time predictive analysis, and service recommendation. Marzieh was also familiar with applications in process engineering to give engineers better insight to improve the process maintenance, safety, reliability and performance. Her experience as an Associate has increased her knowledge as well as her familiarity with other data science opportunities in these and other sectors.
Milan Kordic, AltaML Associate ML Developer, holds a Master of Applied Science (M.A.Sc.) in Electrical and Computer Engineering, and notes that his awareness of data science opportunities in traditional sectors before joining the Applied AI Lab was limited. Milan was first exposed to data science while working for an electric utility company; however, the application of artificial intelligence (AI) and machine learning (ML) was still in its infancy. Milan commented that in those days, “the perception within the traditional sector was that data science was siloed with IT operations and that advanced data science activities such as AI and ML were restricted to academic institutions. Perceptions of data science within the industry started to progress when fast and accessible data analytics platforms were introduced to demonstrate the hidden value available to utilities within their database systems. Traditional companies are now incorporating data science into their strategic initiatives and building dedicated data science teams to transform their businesses.”
Lucas Santos Queiroz, AltaML Associate ML Developer, holds a Master of Engineering (MEng) in Chemical Engineering, and was previously aware of data science applications in finance, banking, and medicine, such as stock market prediction, fraud detection, and disease detection, asking rhetorically “Who has never played with a Kaggle dataset around these topics?” Lucas says that now, the program has opened his eyes to the true potential of machine learning solutions and he sees how data-driven decisions can genuinely impact other industry segments. “For instance,” continued Lucas, “inventory management can be improved by predicting the right amount and product that a customer will need, basically a Just-in-Time system, and another great example is downtime or equipment failure prediction that helps increase the reliability of a chemical process. Machine learning is indeed revolutionizing entire industries, and many applications are being created and developed continuously. The Applied AI Lab is helping us be up to date on the new findings, and also, it is showing how creative thinking can discover new ways to deliver value with ML solutions”.
All three Associates--Marzieh, Milan and Lucas--have backgrounds that align well with Spartan Controls which is advantageous--but, per Milan, such domain expertise is not absolutely necessary for each individual, as long as the team is well balanced: “a strong, well-balanced data science team should have subject matter experts that can leverage their domain expertise to efficiently guide the project lifecycle.” Similarly, Lucas notes that “having domain knowledge can improve the project's success rate by accurately communicating with the industry partner and collecting insights from the correct analysis.” Moreover, machine learning projects are not only about statistical models and data visualization. Lucas continues that “communication is a big part of it as well, and domain knowledge can help the team to be familiar with the industry's jargon and deliver a message to your audience correctly. In addition, when it comes to exploratory data analysis, it is not a trial-and-error process. It is focused on bringing valuable insights into the data. Then, domain knowledge plays a significant role.”
The Associates have all found that their horizons have been broadened as a result of their work in the Applied AI Lab with Spartan Controls. Marzieh emphasized the value of working with real world large datasets and growing familiarity with industry challenges and several new aspects in data science. Milan has learned that not all data science jobs--e.g. ‘Data Engineer’, ‘Machine Learning Engineer’, ‘DevOps’, ‘Analyst’, ‘Data Scientist’-- are the same; in fact, he realized that not all data science jobs are even technical. Milan continued, “having a strong background in mathematics, statistics, and computer programming are staples for being a successful data scientist; however, having strong communication skills, story-telling ability, and presentation skills can also lead you to a career path as a data scientist.” Lucas also emphasized communication, commenting that “previously, I believed that these jobs were mainly focused on backend development, requiring a high level of coding and programming skills. Indeed, programming is essential as a Machine Learning developer, but communication also is. Whether to an internal or external audience, being able to describe how you came to a conclusion, rationally justify your approach, communicate your concerns, or even understanding the client’s needs is similarly valuable. Additionally, a game-changer for me was to see many different companies willing to build a data science team, especially non-tech enterprises. Before, I was aware of only tech companies using machine learning capabilities, and now I am happy to see this shift.”
In terms of advice to those contemplating a career in data science, the Associates emphasized hands-on work including projects with open-source datasets, building on your domain expertise or a learning path that interests you, and building connections. Marzieh recommends joining internship programs such as AltaML’s Applied AI Lab to work with real world data, to become familiar with different challenges in the industry and to be in contact with experts in the field.
Milan noted that data science is industry agnostic and thus opportunistic, so if you have domain expertise then learning data science can present you with opportunities in your own field or if you are a data science expert then learning about a new industry can stimulate your creativity. Milan advises building a portfolio of projects, which “do not have to be technically advanced or even use the most cutting-edge machine learning algorithms; instead, they should be focused on solving a useful problem, demonstrating your thought process, and effectively communicating your results to a broader audience. This will highlight your passion, technical aptitude, and communication skills. Start by selecting a topic that you find compelling and search for open-source datasets that you can experiment with – Kaggle is a great resource for this. Next, organize your project in a public repository using standardized coding practices – GitHub is the perfect tool. Finally, write about your experience, challenges, and lessons learned in a personal blog that you can share with the public – I would recommend LinkedIn or Medium. Best of luck with your career journey!”
Lucas recommends “firstly, if you haven’t started a project yet, just do it. Regardless of the fanciness of the applied statistical model or the unique ML application you might discover, just do it – even a simple linear regression can count! Start making small bets - take a common dataset in Kaggle/UCI and play with it. Then, try to build an online blog where you can show your projects – clearly presents every step of your logical thinking and guide us to your conclusions. At the end of the day, you will exhibit great programming and communication skills at the same time!
Secondly, data science is a vast field and continuously growing, so do not scare yourself. Instead, make the learning path as something you would tap dance to do every day. Be aware of standard machine learning techniques and build upon them.
Thirdly, if you are really into a data science career, do not give up. On the way, there will be several moments that you will find yourself minimized by not attending certain requirements of your desired job or even if your model is not achieving the desired performance. However, you are not alone, and you need to be persistent – it is a marathon, not a sprint.
Finally, connect to people on the field and listen to them. Be humble and ask for advice, tips, and resources. Of course, some people will not answer your message. However, the ones who do will add high value to your career path.”
The last word goes to Lucas who described his experience as an Associate as an incredible career milestone that not only reshaped his career options, but also gave him the opportunity to meet amazing people, create a great network of friends, and have fun during the entire program. Thanks Lucas! It’s been great having you, and we look forward to seeing all of you succeed in the next phases of your careers!
About the Applied AI Lab
The Applied AI Lab was created to accelerate applied AI development through a machine learning and data science internship. The AI Lab is a collaborative industry-led initiative, partnering with ATB, Suncor Energy, Spartan Control, and TransAlta. It helps bridge the talent gap between academic and applied skills by providing interns with guidance and access to real business problems, data, and industry partners. Over the 3-year program term approximately 240 individuals will have the opportunity to gain hands-on experience in applying AI.
For more information contact:
Danielle Gifford, Senior Manager of the Applied AI Lab
