How to build an AI-literate workforce
Jeanne-Louise Viljoen, Data Engineer at PBT Group
As a data engineer working in a space where AI is advancing rapidly, I see daily how the skills landscape is shifting. Companies eager to leverage AI often start by saying they need ‘AI skills,’ but what they need is a workforce that understands both the technology and the business goals it serves. This is about building teams that are not only technically capable but fully AI-literate.
One of the biggest challenges I encounter is that businesses aren’t always clear on what they want from AI. They may know they want to adopt it but often aren’t sure what kind of models they need or what data will be required to support them. Without that upfront clarity, even highly skilled teams can struggle to deliver meaningful results. That’s why domain specialists and product owners play a critical role from the outset. These individuals help define scope and ensure that the AI solution is grounded in real business needs.
The evolving role of data professionals
At the same time, I’ve seen first-hand how the way we work as data professionals is evolving. My own role has shifted from traditional data engineering to something much closer to software development. We’re coding more, building complex data pipelines, automating quality checks, and versioning data products as standard practice.
Agile methodologies, continuous integration, and continuous delivery are now part of our daily work. Data itself has become a product. With real-time streaming data constantly updating, we need to move faster and more flexibly than ever. Development cycles that once took months now happen in weeks.
What makes an AI-ready team
To build an effective AI-ready team, organisations need a range of expertise.
They need domain specialists who understand the business, alongside project managers who can bridge business needs with technical teams. Data architects and modellers design how data is structured, stored, and secured, while BI developers and analysts explore and visualise data to generate insights.
Data engineers like me prepare reliable, usable data for AI models, while data scientists take that data and train, refine, and optimise the models. Machine learning engineers maintain and deploy those models into production environments, ensuring they perform as intended.
Testers play a crucial role by validating not only data pipelines, but also assessing model fairness, robustness, and security before solutions are deployed. Finally, governance and compliance specialists ensure that privacy, ethics, and regulations are upheld throughout every stage of the data lifecycle.
AI literacy for every role
What’s become increasingly clear is that every role now needs some working knowledge of AI. Even testers need specialised training to evaluate bias, fairness, and security in AI models. Business teams, too, need to understand what they’re asking AI to do. Ultimately, the models are only as good as the data and objectives they’re built on.
Continuous learning built into the day
Continuous learning is non-negotiable. The people who struggle most are often those who resist learning new tools or approaches. Adaptability is critical. I’ve also seen how powerful it can be when organisations make learning part of the workday. Short, focused training sessions, for example, 15-minute knowledge drops built into project cycles, help teams stay current without disrupting delivery. This kind of embedded upskilling keeps everyone sharp as technologies evolve.
The real key to unlocking AI’s value
Building an AI-literate workforce isn’t just about hiring more data scientists. It’s about creating an environment where everyone, from business to technical teams, understands AI’s role, works collaboratively, and stays curious as the technology continues to advance. That’s how we fully unlock its value.
