Aparna Pratap holds a pivotal position at Mercedes Benz Research and Development India (MBRDI), where she spearheads groundbreaking research and development efforts with a strong data strategy focused on enhancing driver safety and comfort through innovative AI-driven systems. She currently leads a multi-disciplinary AI Platforms team from India. This is her AI journey.

Can you tell us about your AI journey?  

I started my career in aerospace, where I worked for nearly ten years. After my maternity break, when I was ready to rejoin the workforce, I found an interesting opportunity in data analysis focused on large aviation engine data. That was my introduction to data analytics and prediction models. It was also the inflexion point for deep learning, Convolutional Neural Networks (CNNs) and their successful use in AI applications and commercialisation. That's when I joined Mercedes-Benz Research and Development India (MBRDI).  

Driven by my interest in understanding this space better, I learnt statistical modelling, a technology stack that helps handle large-scale data. This learning experience helped me in my first assignment at MBRDI, which involved establishing a team that delivered large-scale image datasets to train vision-based deep learning algorithms in the automotive domain. We then focused on the evolution of Machine Learning Operations (ML Ops) and the design of scalable systems that can industrialise AI-based solutions. Currently, I lead the AI Platforms team from India. 

What is your area of expertise in AI, and what made you choose it? 

I collaborate with multi-functional teams, building AI platforms for production-grade systems. I specialise in computer vision and image processing, where I have gained significant knowledge and experience. Looking ahead, I believe that the future of large-scale computing will be focused on quantum technology.  

How important do you think skilling and upskilling are in a country like India?  

India's young and vibrant demography is a tremendous asset for the country's growth and development. Continuous skill enhancement in the tech sector is crucial for talent to thrive in today's rapidly evolving digital world. With immense potential for innovation in engineering and technology, there has never been a better time to be an engineer. 

Strong expertise in engineering-intensive areas such as control systems, product and design thinking, and fundamental sciences like Math and Physics will be essential for building solutions for the future, be it in autonomous driving or fundamental biology like protein folding. On the technology front, significant research areas include advanced AI with reasoning capabilities (Artificial General Intelligence), multi-modal AI, AI agents, generative AI, sensor fusion, and statistical safety validation. Another important area is compressing models using quantisation methods to deliver personalised solutions on the edge. These advancements will significantly impact how we build and design real-world applications today. 

As science and technology undergo advancements, our ability to adapt is key. Fostering first-principle-based thinking at different levels is essential for navigating these changes effectively. At an individual level, empathy, teamwork, effective communication, an appetite for risk, and the ability to collaborate in global teams are critical. 

Throughout my career, I have pursued continuous learning, taking academic courses in mathematics for machine learning and convex optimisation to stay connected with the foundational knowledge necessary to understand the theoretical underpinnings of machine learning algorithms and optimization techniques. 

Describe some challenges you have faced in reaching where you are now. 

Pivoting to an entirely new technology landscape and way of working after returning from a break has been my greatest challenge. Although I managed to stay updated and upskilled during my break, I missed the industry connection. It took extra effort to take on industry-relevant projects and assignments during this time. Mentorship and acceptance of the workforce that continues to work on industry-relevant assignments in preparation for full-time engagement is important to bridge this gap. 

In what direction should companies in the AI sphere move to ensure more female participation and leadership?   

We must encourage more girls and women to pursue careers in STEM. There is still a significant dropout rate of girls from STEM subjects after schooling. Building aspiration by bringing in more relatable role models for girls to look up to is important. At work, it is equally important to encourage women to take leadership positions, giving them visibility and recognition. Women must also advocate for themselves, actively negotiating at the table and for the larger workforce they influence. Alongside hard work, we must embrace self-care and promote personal well-being.  

In a country like India, where societal constraints bind women, how do you think that they should break the stereotypes? What are your words of wisdom to young girls who wish to build careers in AI?  

Speak up, negotiate, show up, network, influence, question the status quo, and invest in regularly upskilling yourself.  Careers in AI can be diverse, from applying AI to enhance creativity and productivity to building large AI systems that drive the next generation of innovation. We must approach our career by becoming experts in fields such as engineering, arts, design, finance and then complementing this with AI competence in specific areas.