The World Economic Forum Global Gender Gap Report 2021 states that women take up only 32% of the total workforce in data and AI, 20% in engineering and 14% in cloud computing. Many times, we see young talented women being discouraged from taking up STEM-based studies. A great way to motivate young girls to enter this space is by reading stories about women who have broken the glass ceiling and taken up challenging roles like data scientists, AI practitioners, and machine learning engineers.
Today, we look at the inspiring journey of Swetha G Basavaraj, Director of Product Management, Data and AI Samsung Electronics, America. Basavaraj comes with extensive experience in major companies such as Yahoo, Volvo Cars, DataVisor and IBM. She has also built her own company, SAPling Software Solutions.
Basavaraj feels that critical thinking and problem solving were some of the prerequisites for her for any choice of career. She said, “I have spent many years as an engineer and became an entrepreneur trying to solve a business problem using technology solutions. At that time, I didn’t realise I was more or less acting as a product manager. After my stint as an entrepreneur, I went to do my Masters at Stanford GSB, where I spent time learning Design and Digital Advertising domains. After my Masters, I was looking for a role where I could solve large-user-base problems combining business and technology. Having enjoyed entrepreneurship in my past, product management seemed like a good next career step. I joined Yahoo as a product manager for their Ad-tech platform.”
Subscribe to our Newsletter
Join our editors every weekday evening as they steer you through the most significant news of the day, introduce you to fresh perspectives, and provide unexpected moments of joy
AI and analytics
While working as an engineer, Basavaraj had some experience in business intelligence tools and analytics. When she joined Yahoo, the scale of data and the technology to utilise and monetise that data was fascinating. She made the deliberate career move soon after to work more closely with AI, and that led her to DataVisor, which has been recognised for multiple years as one of the top AI startups to look out for in recent years.
She added, “At Samsung, I manage a global team of product managers who help identify and define business problems that can be solved using data and machine learning. My team also manages a global data platform within Samsung that can be utilised by analytics and machine learning teams to achieve those business goals.”
In the long and fulfilling career that Basavaraj has had, she has encountered quite a few challenges. The most crucial of them were:
- Products involving data and AI are very technical, so it is very easy to get lost in the details. We have to constantly remind ourselves and the team about the business outcomes that have to be met. Since the outcome is performance-centric, engineers and data scientists on the ground overlook data privacy, security, and governance (Data and ML). So, educating them, creating a process and maintaining this on a regular basis is paramount to building responsible AI and sustainable data strategy.
- It is easy to build a proof of concept but building a scalable long-lasting solution takes planning, effort and budget. Many ML models after successful proof of concept never make it to production because of both gaps in architecture required for such a deployment as well as lack of collaboration between data, IT and ML teams. So, one needs to have DevOps and MLOps as part of the overall planning.
- Maintaining a feedback loop to keep improving the quality and adoption after hitting a certain size – not just in technology, where we evaluate the models and fine-tune them but also as a product manager and lead evaluating business metrics and redefining the problem statement. Most importantly, hiring for the right skills within the team is crucial for team excellence in all of the above.
Master the fundamentals to solve the right problems
The only constant in the technology field is change. Basavaraj added, “As we are collecting more and more data, we are increasingly relying on greater automation and faster experimentation. So, it becomes imperative to leverage advances in technology to keep up with the market demand. 80% of the data we see has been created in less than five years, and so, to make sense of this data, we need updated technology solutions.”
As a data science professional, one should not only keep themselves updated with the latest technology solutions but also master the first principles/fundamentals to solve the right problems. As data grows, having good data engineering skills also becomes increasingly important.
Basavaraj stated, “My only suggestion to the folks considering AI as a profession is to assess yourself first on your interests before just following the trend. There is so much information on the internet and many channels like Analytics India Magazine, which provide you with a forum to both learn and explore Data science as a career that you should take advantage of.”
Traits of a good data scientist
Typically for any data science/AI role, there are at least three different areas Swetha considers:
- Critical thinking and ability to learn/adapt to new problem statements or challenges
- Technology know-how – Working with large scale (volume, variety, velocity) data and new breakthrough algorithms/solutions in AI to solve data. Nowadays, data scientists can also grow more self-sufficient with data engineering skills to unblock themselves and reduce dependency. So it is important to be familiar with data engineering tools and technology for faster iterations.
- Domain expertise – Someone who is either passionate or has experience in that domain. It is a combination of skills coupled with curiosity to learn new paradigms.
Start small and go deep
If a college student or someone who is a freshly passed out aspirant plans to pursue a career in data science and AI, Basavaraj feels the following points should be kept in mind.
- Build your business problem solving and technical skills by working on hands-on projects. You have potentially four tracks: ML engineer, Data Analyst, Data Scientist and Data Engineer, and each requires different types of skills, so know your strengths and choose the right function.
- Technology and algorithms are ever-changing, so don’t stress over the vastness of the field. Start small and go deep. Fall in love with the problems and find technologies to solve those problems.
Break this fixed image of women being good only in certain types of roles
The diversity crisis in data science/AI is real. Basavaraj feels, “More broadly, in general, the technology industry is still dominated by men with less than ~30% of the workforce being women. So, the scarcity of women in AI, which requires deeper technology skills, is not surprising.”
Basavaraj says that there is a need for more women to choose STEM and continue to pursue career growth in technology. We should break this oversimplified and fixed image of women being good only in certain types of roles. Business leaders need to consciously support them with transitions to changing technologies in AI and give them a platform to grow within an organisation.
“Women are key to scaling up AI as a practice, and the onus is on both sides – demand and supply. We want women to step up, creating enough workforce for organisations to choose from and organisations to open up enough opportunities that can be fulfilled by women. If you look at the stats in India, only 33 per cent choose STEM. So, we need to catch up young and have female role models to promote young girls in STEM careers. Once in the tech ecosystem, we need advocates to encourage and support women to train and build that pipeline of potential AI or data science candidates and establish equal employment opportunity”, Basavaraj adds.
Data scientists becoming part of the core product teams
We are in a fascinating stage for AI and analytics in the tech industry. Just a few years ago, AI was treated as a research project or was worked on in silos, but we are now seeing data scientists becoming part of the core product teams driving business outcomes in production at scale.
Basavaraj concludes, “As we see ML/AI model results becoming as good as or sometimes better than humans, we will see more automation, and this technology will get embedded in almost all walks of life. We will also see more specific areas of expertise within AI, such as cybersecurity, NLP, computer vision, etc., as this field advances. Model and Data Governance will become a necessary function within data and ML teams as we see organisations making responsible AI part of their core value.”