How Yoga, Asian Business, and Stanford Helped this ML Engineer Get a Job at Meta Meta's ML engineer Aleksandr Timashov talks about his journey from math classes in a small Russian town to ML engineer position at Meta

Aleksandr Timashov fell in love with numbers during his childhood. At the age of 13, he joined a math lyceum in his small Russian town Yefremov. 'That school was one of the most innovative in the city, providing each student with a brand-new desk and new books, which was quite unusual for Russian schools at that time', he recalls.

This environment helped him to develop better understanding of math, win Olympiads and, eventually, enter the prestigious Lomonosov Moscow State University, to pursue a degree in Mechanics and Mathematics. He soon, however, found the program too theoretical for his tastes and withdrew after the third year, and later completed his studies at Indiana University East, located in Richmond, USA, from which he eventually graduated.

His academic journey shaped him as a skilled specialist in machine learning (ML) with ambitions for a high-powered career. He started his journey as a young professional from creating scoring algorithms that enhanced decision-making and reduced risks for one of the prominent Russian banks in 2012. He then moved to freelance work in various industries and ended up in Asia. In 2018, he landed a position of machine learning engineer in a new sector for him - fashion.

The job was with a Malaysian-based company called Omnilytics, which gathers data for fashion brands to help them make better business decisions. He mentioned that he found the job unexpectedly while in Malaysia: 'I went to yoga, and when the yoga teacher learned that I was an engineer, she mentioned that a friend was looking for an ML engineer for a startup. We met pretty quickly, and even though my English was at a zero level, I was offered the job.'

During his time at Omnilytics from 2018 to 2019, his main task was to develop matching algorithms capable of comparing 10 million products in real-time. It was based on computer vision (CV) and natural language processing (NLP) models that he trained. Reflecting on this experience, he commented, 'That was a crucial period for me. Creating large-scale feature comparison models provided me with practical insights into integrating deep learning with real business needs – skills that would go on to shape my career.'

After a short period at Entropia Global, Timashov was hired by Petronas Bhd in 2019, Malaysia's oil and gas giant, as the executive in data science and was quickly promoted to the head of the department and lead machine learning Engineer. His main focus was on applying machine learning in computer vision, and his contributions there were truly transformative.

One of Timashov's most significant achievements at Petronas was optimizing the company's security manpower by 30 per cent, thanks to a real-time video-analytic system he helped design and implement. This system utilized convolutional neural networks (CNNs) for person detection, re-identification, and tracking—core applications of AI in industrial security and surveillance. He also improved social distancing compliance by applying graph traversal algorithms, contributing to public safety during the COVID-19 pandemic.

Timashov's innovations not only improved safety and security at Petronas, but they also saved the company over $10 million by automating labor-intensive industrial plant inspections. This is a critical example of how AI can bring significant cost savings and efficiency improvements to traditionally manual processes.

In addition to his technical accomplishments, Timashov is passionate about leadership and mentorship. At Petronas, he mentored and coached more than 10 machine learning engineers, fostering a culture of curiosity, continuous learning, and collaboration. He believes that people are at their most productive when they are engaged in work they love, and to nurture this, he hires individuals who are naturally curious and driven to explore new technologies and ideas.

Timashov encourages a work environment based on trust, open communication, and collaboration. One of his most effective techniques for building strong teams is organizing members into small workgroups or pairs. This setup fosters collaboration, blends diverse skill sets, and helps solve complex problems by leveraging the collective strength of the team. 'This team-focused strategy has been key to pushing the boundaries of innovation throughout my career,' he adds.

Even though he achieved a lot during his time at Petronas, in 2021 he realized that he wanted to move to the next level. 'I need to look up to and learn from people, but at Petronas, I was the strongest, and people learned from me,' he explains.

Therefore, in 2021, he enrolled in a program in artificial intelligence at Stanford University. That was a fruitful decision, he believes. Not only did he learn more about deep generative models, natural language processing, etc., he also sparked interest from some major IT companies in the world, such as Tesla, Google, and Meta. 'Stanford was a green light for them. It was obvious from the message of Tesla, for example. The company wrote directly like this: 'We see that you studied at Stanford, so we want to communicate with you.' Finally, in October 2022, he accepted work as an ML software engineer at Meta in the London office. In this role, he continues to lead innovative projects at the intersection of AI, computer vision, and real-time data processing.

He believes that individuals who grew up in even the smallest towns in the world can follow his steps and find employment with major global companies. The key is continuous learning, adaptability across different industries and geographies, and a bit of luck.