Karl Crowther discusses how orchestration, human accountability and workforce upskilling can help organisations scale agentic automation
Gulf enterprises are moving beyond AI experimentation, with attention shifting towards measurable business value, governance and the skills needed to scale adoption. Clear objectives and coordinated workflows are becoming increasingly important as organisations bring AI agents, software robots and people together across business operations.
Speaking to TahawulTech.com during UiPath FUSION Dubai on Monday Karl Crowther, Vice President, Middle East, Africa, India and South East Asia at UiPath, discusses the differing stages of AI maturity in the UAE and Saudi Arabia. He explains why successful deployments depend on identifying high-value processes, maintaining human accountability and equipping employees to work effectively with AI.
Interview excerpts
How do enterprise AI maturity and customer demand differ between the UAE and Saudi Arabia?
Both countries have moved beyond asking whether they should adopt AI. Both are now doing it. I would say the UAE is further along the maturity curve, while Saudi Arabia is investing heavily to catch up, particularly through Vision 2030. From an automation, platform deployment and scalability perspective, the UAE is scaling, while Saudi Arabia is putting platforms in place to enable that scale. Much of Saudi Arabia’s momentum is being driven by the ambitions of Vision 2030.
What prevents Gulf enterprises from turning AI pilots into measurable business outcomes, and how can they overcome those barriers?
Over the past couple of years, some organisations were running pilots for the sake of running pilots. The focus was often on responding to a mandate to adopt AI or agentic technology, without fully considering the underlying technology, data and processes. Consequently, quite a few pilots did not necessarily deliver successful business outcomes. We are now seeing a more considered approach across sectors such as banking, government and aviation. Organisations are asking what a successful outcome looks like, how to define it and which high-value processes could deliver a meaningful return on investment. This is partly a question of maturity. The initial hype encouraged experimentation, but organisations are now pausing to think about the value they want to achieve. That clearer focus on outcomes is making pilots more successful, in my view.
How are AI agents, robots and people working together to automate complex operations?
The key is orchestration. UiPath provides a platform to orchestrate agents, human involvement and the different processes required to deliver an end-to-end outcome. Some organisations in the UAE already have agentic use cases operating across an entire process, with humans remaining involved to provide ownership and accountability. Agents perform high-value tasks within that process, while people retain responsibility for the outcome. The value comes from bringing those elements together into a complete workflow. Individual agents need to operate within the context of the wider process, with clear roles for both technology and people.
How should enterprises assess the costs and productivity gains of agentic automation before expanding deployments?
The starting point is the usual assessment of total cost of ownership and return on investment. However, different projects are driven by different objectives. Cost is often one consideration, but the desired outcome could also involve revenue growth, profitability, productivity or reducing the staff time required for particular tasks. Organisations need to define the outcome first and establish what successful delivery would look like. The measures will differ by sector. A public sector organisation might focus on maintaining or improving services for the population. A bank’s risk function might want to reduce false positives in an anti-money laundering system.
The essential step is to define the business outcome, attach measurable indicators to it and ensure the project delivers against those expectations.
Where should enterprises draw the line between autonomous decisions and human oversight, and how does UiPath help customers enforce those boundaries?
This is a difficult question because the technology is evolving so quickly. There is also a deeper philosophical question about who is responsible for AI and how human accountability should be incorporated. Regulation, government policy, local rules, data sovereignty, and data residency all influence those decisions. Whatever the context, there has to be accountability, supported by a platform that enables governance and security. An enterprise can deploy many agents, but without accountability, security and governance, how does it control the outcome? From an orchestration perspective, the UiPath platform allows organisations to coordinate people and agents within a governed process. The boundaries continue to evolve, but governance and human accountability must remain central to how these systems operate.
How can organisations equip their workforce to use AI effectively while maintaining the skills needed to deliver results?
Every organisation is working through this, including UiPath in our own internal use of AI. Much of it comes down to workforce enablement. Organisations need to choose the platforms they want to adopt and ensure their people are equipped to use them. Upskilling is particularly important and can be one of the major benefits of AI adoption. The principle is similar to robotic process automation. When technology supplements someone’s role effectively, that person can spend more time on high-value work and less time on repetitive, routine tasks. AI creates a similar opportunity. Employees need to understand how their roles are changing, what now constitutes high-value work and how AI can support them as a copilot. My strongest advice is to prioritise enablement and upskilling so that people can use these tools effectively.
Source: Tahawul Tech