A finance associate’s race to become a PETRONAS machine learning champion

by Training and Certification Blog Editor | on

How PETRONAS’ Amazon Web Services Skills Guild propelled Aiyna Nazieha Muhamad to become the first female PETRONAS Exclusive Amazon Web Services Virtual DeepRacer winner — with the fastest time ever recorded in Malaysia.

Since 2019, PETRONAS , Malaysia’s national petroleum company, has been going through an extensive digital transformation, embedding data, analytics, and automation into almost every aspect of the organization. While moving the majority of their workloads to the cloud, they also created PETRONAS Digital Academy in 2021 to invest and equip their employees with the necessary technical skills. Two years forward, in mid-2023, a PETRONAS Exclusive Amazon Web Services Virtual DeepRacer competition was organized in conjunction with PETRONAS Digital Day 2023.

The competition allowed the employees to put their newly acquired skills to the ultimate test but it also proved a glimpse of how PETRONAS Digital Academy was working and how PETRONAS was a step closer to achieving ‘Digital Democracy’ where everybody can affect change through effective use of data and analytics.

Initially designed as a fun way to improve people’s analytics skills and abilities, Amazon Web Services DeepRacer has evolved into a worldwide league where racers compete for glory and cash prizes by creating and training a machine learning model to control a ⅛ scale autonomous race car around a track in the fastest time possible. Given the strong digital backbone at PETRONAS, there was little doubt there would be significant talent participating in their PETRONAS Exclusive Amazon Web Services Virtual DeepRacer event. However, it wasn’t any of the company’s digital professionals who won the grand prize, but Aiyna Nazieha Muhamad, a controls and compliance associate who works in PETRONAS’s finance department. She not only won the competition as the first female winner but also with the shortest lap time in Malaysia.

Aiyna Nazieha Muhamad

Aiyna’s win is a testament to how deeply PETRONAS employees are steeped in the culture of data, analytics, and digital technologies thanks to Tan Sri Tengku Muhammad Taufik, President, and Group Chief Executive Officer of PETRONAS for championing a digitally fluent workforce enterprise-wide. This is supported by PETRONAS Digital Academy, which is 100% focused on upskilling and including as many employees as possible in their training efforts. In fact, one of the first things the Digital Academy pushed for HR to create was what is known as Learning FriYays, where after lunch on Fridays, PETRONAS employees are encouraged to spend their working hours learning and upskilling themselves via the internal learning platforms.

From the beginning, the PETRONAS Digital Academy partnered with the Amazon Web Services Training and Certification to make a wide range of Amazon Web Services Training opportunities available. Using instructor-led training sessions, hands-on learning sessions, and the self-paced digital learning platform Amazon Web Services Skill Builder, more than 500 PETRONAS employees have learned new Amazon Web Services Cloud skills. Many have gone on to earn Amazon Web Services Certifications in various disciplines to validate their technical skills and cloud expertise.

Using Amazon Web Services Skills Guild framework, PETRONAS’s Digital Academy focused on cohort-based training to scale upskilling. This uplifted technical skills in cohorts and created a network of knowledge within the organization, which led to exponential collaborative effort.

This culture of constant skills development could not have been more suited for someone like Aiyna, the winner of PETRONAS Exclusive Amazon Web Services DeepRacer Competition. Although all her formal studies were in finance, she has long been interested in technology and machine learning. “Before enrolling in the training at PETRONAS, I considered myself a self-taught coder with a skill level of two out of five,” she says. “My machine learning skills were even more basic. I spent a lot of time learning the theoretical underpinnings, but because I wasn’t applying it day to day, I had a hard time imagining how it worked in practice.”

Digital Academy organized a day of virtual instructor lead training on machine learning by our Data Science team. And the DeepRacer competition and access to training resources through Amazon Web Services Skills Guild proved a powerful force multiplier for her. Amazon Web Services Skill Builder format also allowed her to learn at her own pace and on her schedule. “I was so excited to learn about machine learning that it was easy for me to make time for it,” Aiyna recalls. “I was really starting from scratch, so I actually went through the  Amazon Web Services DeepRacer machine learning course  available through Amazon Web Services Skill Builder three times to refresh my training and to make sure I understood everything fully, taking it in 30-minute sessions. The tests along the way helped as well.”

Aiyna’s initial attempts at training her models were inauspicious. The first few she trained couldn’t even stay on the track, let alone finish with a competitive time. “After several hours of crashing, I threw out all the models I had trained and started over with the simplest code I could write,” she says. “After only two or three tries, that model was able to stay on the track. That gave me confidence that I could at least finish the race. After that, though, I dropped any idea of being competitive with the other racers. I was just competing with myself.”

To her relief, the contest turned out to be more collaborative than cutthroat, a testament to the culture of learning PETRONAS has established among its employees. “There was a group chat where people in the course could discuss concepts and share information,” Aiyna says. “It was wonderful. Everyone was so committed, and also generous with their knowledge.” Ultimately, after several days of tweaking and training, she crossed her fingers and submitted the best model she was able to come up with.

A few days later Aiyna was amazed to hear that her DeepRacer had finished in the top 6. She attended the award ceremony, eager to see whose model had won, and how much they had beaten her by. She could not have been more shocked to hear her name called the grand prize winner. Not only was she the first woman to ever win the DeepRacer contest at PETRONAS, but her lap time of 9.801 seconds was the best ever recorded in Malaysia. “I was super confused and excited at the same time,” she says. “It was the most amazing experience of my life.”

Aiyna’s contest success brought more than a day of thrill. For Aiyna, all her personal work learning to code, working with big data, and getting hands-on experience with machine learning has significantly broadened her understanding of the role of cloud analytics for PETRONAS’s finance department, as well as giving her a community of peers she can contact for advice or assistance. “The connections I made through taking on this project have been invaluable,” she says. “The cloud has a great deal of applicability for what we do in finance. And now, I have a community I can reach out to for support.”

Aiyna is considering some even bigger changes in the future. “I’m interested in further polishing my coding skills, and possibly even pivoting into a data role at some point”. Her ‘give it a go’ attitude and penchant to learn new skills is an inspiration to other Malaysian women contemplating a career in technology. “If you’re a woman thinking about picking up machine learning or trying something like DeepRacer, all I can say is go for it. Don’t be too hard on yourself and don’t give up. It might not be easy, but if something is interesting to you, it’s worth checking it out, even if it’s outside your comfort zone.”

To achieve digital democracy at PETRONAS, PETRONAS Digital Academy is providing diverse learning opportunities and modalities to build a cloud-first culture and enhance digital fluency organization-wide.

Read more about Amazon Web Services Skills Guild