APAC career guide on AI

What an AI career looks like

AI builders and engineers are at the core of many AI initiatives as they build tools across the entire AI stack. For example, those in AI roles at AWS are charged with designing machine learning (ML) algorithms and systems, improving the performance of deep learning models, and working with customers to train and deploy ML models in the cloud and on edge devices at scale.

AI careers can also involve building and administering big data for ML models, assessing complicated datasets to recognise business trends, and developing and refining prompts for generative AI services.

“AI roles are also constantly evolving to keep pace with the development of the technology. What we know of AI careers today may look vastly different tomorrow as AI goes mainstream”

“The need for integrity in data, robust security and a discipline built on engineering fundamentals will grow in importance as the use of AI accelerates. At AWS, our training offerings are uniquely equipped to support individuals seeking a career in this fascinating space,” says Sklar.

Some examples of AI careers, as pointed out by Sklar, include:

• Software development engineers – who design, develop, test and deploy distributed machine learning systems and large-scale solutions for users.
• Applied scientists – who work on systematic approaches to improve the performance of deep learning models.
• Data scientists – who straddle the business and technical worlds with deep data analysis to achieve specific outcomes. In the field of ML, they design and build models from data, create and work on algorithms, and train models to predict and achieve business goals.
• AI engineers – who develop the tools, systems and processes that enable AI models to be applied in the real world.
• AI ethicists – who advise on ethical AI practices and guard against bias, unintended consequences, and ensure accountability within the organisation.
• Prompt engineers – who develop text-based prompts to refine the responses of AI models.
• Product managers – who design and launch new AI products and services to market and provide AI advisory and go-to-market advice, from use case and business case development through to AI ethics, including the responsible use of AI.

Skills and certifications

Gavin Barfield, vice-president and chief technology officer of solutions at Salesforce ASEAN, notes that with AI teams being increasingly cross-functional due to the growing number of real-world AI applications, there are opportunities for engineers to deliver business value from AI by solving real problems and driving results for customers.

To do so, they will not only need technical skills in areas like data science, machine learning, neural networks and prompt engineering, they should also develop industry skills and knowledge so they can collaborate more closely with the business and customers to identify use cases where AI can add real value.

To thrive as an AI engineer, DBS’s Tuomola says a solid educational foundation combined with a diverse skillset is crucial. Typically, AI engineers have a degree in computer science, software engineering, or a related field. In addition, they would need the following skills:

• Proficiency in programming languages like Python, enabling them to implement AI algorithms effectively.
• Expertise in machine learning frameworks such as TensorFlow and PyTorch, empowering them to build advanced AI models and systems.
• Strong problem-solving abilities and analytical skills to join the dots in tackling complex challenges and devise innovative solutions, while taking a customer-centric lens.
• In-depth knowledge of data structures and algorithms, which form the backbone of efficient AI implementations

While each of the above skills is valuable, Tuomola believes an understanding of AI and ML algorithms, and having the software engineering skills to apply them effectively, are paramount.

“Algorithms serve as the foundation of AI models and systems, and a profound understanding of them is essential for designing cutting-edge solutions that best fit our business needs,” he says.

“At the same time, productionising such models requires the application of good software engineering practices, to ensure the system meets its requirements for scalability, performance, availability and maintainability. Bringing these skills together empowers AI engineers to develop innovative approaches, optimise performance and unlock the full potential of AI technologies.”

Besides technical skills, it is also key for AI engineers to have critical thinking and problem-solving skills, business acumen, as well as interpersonal skills such as communication and collaboration, says AWS’s Sklar.

This will enable them to look at multiple variables to determine the best course of action, and work with internal stakeholders and project teams quickly to support their organisation’s business goals, he adds.

Those who are looking to courses to kickstart their AI careers can turn to online materials and courses, such as Stanford University’s ML course on Coursera and the AI programme by the University of Helsinki.

Most cutting-edge research in AI/ML is also published for free in arXiv.org, an open-access repository, making it easy to keep up to date with the latest research trends. To learn practical skills and to test your skills against other AI engineers, Tuomola recommends trying out competitions at Kaggle, an online community for AI practitioners.

There are also AI/ML engineering courses provided by different cloud providers. AWS, for example, offers over 70 free digital training courses covering AWS ML services and solutions, with content ranging from foundational topics such as ML essentials for business and technical decision-makers, to advanced topics such as building ML applications.

“When you think of careers in AI, you probably want to think less about coding, AI research, data science and engineering, and more about consumption”

Budding AI professionals can also consider AWS’s ML Learning Plan designed to help AI engineers integrate AI and ML into tools and applications and prepare for the AWS Certified Machine Learning – Specialty certification examination.

Nvidia, whose GPUs are frequently used to crunch AI workloads, has training programmes to meet the needs of data scientists, application builders, creators, technical artists, researchers and IT administrators, covering the skills needed to architect, build, deploy, operate and support AI-based solutions and platforms.

But how important are certifications in one’s AI career? Salesforce’s Barfield contends that AI professionals should focus on the proficiencies they’ve gained through certifications, rather than the certification itself. “Gaining industry knowledge is also important, as it helps AI professionals better contextualise and understand the real-world applications and benefits of AI,” he adds.

Tuomola notes that certifications should not be seen as a substitute for practical experience. Instead, real-world experience – working on challenging projects and solving complex business problems – is the most valuable asset one can possess in the field of AI. “While certifications provide a foundation and enhance your skills, it is your ability to innovate, think critically, and apply your knowledge in practical settings that sets you apart.”

Making AI consumable

With AI capabilities increasingly being democratised through cloud services, the AI skills in greatest demand may not necessarily be technical in nature, contends Allan Waddell, co-CEO of Kablamo, an Australian technology services firm.

“When you think of careers in AI, you probably want to think less about coding, AI research, data science and engineering, and more about consumption,” says Waddell, who has a computing degree. “If I was doing my degree again, and doing AI, I’d be thinking double degrees in law and ethics. I’d be really resistant to going into a siloed AI degree – I’d maybe do a post-grad in it.”

And he explains why: “There won’t be a huge need for AI technology specialists, but there will be a big need for harnessing how it is applied. There are going to be very few people working on the optimisation of AI, especially when AI starts doing this optimisation itself, so I’d be looking at the application of AI.

“I’d be looking at digital product development using AI, making the complex simple. Consumption of AI is what matters now. How do you make AI consumable to your customers? It might be that user experience might be more important now than anything else – build the bridge from generative AI to what a business actually needs.”

On future-proofing one’s AI career, Waddell does not think it’s about future-proofing the current roles in AI, especially those that are highly specific and technical in nature. “It’s about how you stay general enough so you can pivot. You don’t want to get so specific now that there’s no way to get out.”

“Currently, AI is basically just another opinion, a probability engine generating an outcome. Gross replacement of careers isn’t happening yet and probably won’t for a while – we’re in a hype cycle around what AI is imminently capable of.

“Think of self-driving cars and all the money that’s been poured into that – there’s an almost infinite number of edge cases you have to train for, and we are still far from getting it right. The biggest risk around AI – and any specific careers in it – is the threat of premature regulation that stifles its growth and limits our options.”