Agentic Success Requires These Vital Skills – and They’re Not All Technical

As supply chain and operations leaders grapple with AI developments, the roles and broader skills required for implementation are shifting. In some ways, this shift is unsurprising: Our latest analysis found that 22.1% of companies are now hiring for multiple skills relating to agentic AI in supply chain roles. This has extended to the executive level, with a 428% increase in mentions of AI skills in roles for C-suite positions between 2022-2024.  

But what’s noteworthy is the shift happening within the realm of non-technical skills. AI might be expected to replace some white collar jobs, but it’s also proving some more “human” skills to be distinctly... well, human.  

AI Skills at the Helm? 

Delving deeper into the specifics of AI-heavy roles, certain positions are surging onto the supply chain scene. While job posts mentioning agentic AI have increased 2.5x since 2022, hiring for data scientist/knowledge engineers has doubled and posts for semantic technology developers have grown 5x. But what do these folks actually do? 

Let's start with what most companies are trying to achieve: launching agents that can deliver an outcome. This requires agents to ingest signals and process them against business context to recommend or even execute a decision. But to get that right, we have to feed them the business knowledge to reason against.  

Agents get this knowledge and context from the semantic layer, where data becomes understandable to a machine. LLMs (which are trained to find patterns in massive unstructured text and image data sets) and knowledge graphs (which create explicit relationships between data entities) are companions to that semantic layer, meaning skilled talent – like knowledge engineers – who can design and engineer the data architecture necessary for agents are in high demand. No wonder these roles are surging – in a recent Zero100 survey, 46% of respondents said data issues were a barrier to AI implementation.  

And even if we aren’t involved in constructing AI, the AI literacy bar is rising. Industry giants are addressing this through upskilling – in the US, Microsoft has committed $4 billion to “ensure that students in every school across the country have access to AI education,” while Amazon has partnered with Udacity to deliver fully funded nano degrees under the AI/ML Scholars program to train future AI scientists, business intelligence engineers, and AI engineers.  

Human Skills at the Helm? 

While many roles, skills, and companies have a clear bias toward tech, there’s another side of the coin. AI is eliminating some roles, but the limits of machines are being tested. Simply using AI everywhere just doesn’t guarantee a slam-dunk experience for users. After all, the top use of ChatGPT in 2025 might have been therapy – but it doesn’t mean it outperforms a human therapist.  

Fully understanding the relationship between digital and human ecosystems is very much still a work in process, evidenced by Digital Humanities degree offerings sprouting up at institutions like Cambridge University, where the impact of the digital world and its effect on human systems are studied.  

In practice, this highlights the necessity of humanity in the workforce. A study by the Organisation for Economic Co-operation and Development (OECD) shows that in roles with high AI exposure, there has been an increase in demand for non-tech capabilities such as social skills, attitudes (or “emotional skills”), languages, and cognitive skills. Even more interesting is that the most in-demand skill in high AI exposure environments is originality. This includes creativity and developing new ideas, both of which saw an average increase in demand from 25% and 33% between base and end years.  

New Capabilities, New Relationships 

There is no doubt that we are marching toward a future where we work shoulder to shoulder with agents and need to develop and build new technical skills to harness all they have to offer. But as supply chain evolves, our uniquely human ingenuity will continue to have a solid role to play. After all, as the World Economic Forum puts it, “AI must serve human creativity, not replace it.”