As enterprises accelerate AI adoption, many organisations are discovering that technology alone is not enough to deliver meaningful outcomes. Fragmented business processes, disconnected data and organisational silos continue to limit AI’s ability to scale across the enterprise. In this exclusive interaction with Tech Achieve Media (TAM), Manuel Haug, Field CTO, Celonis discusses why operational readiness is becoming the foundation of successful AI initiatives, how enterprises can move beyond isolated AI use cases, the evolving role of AI agents alongside human teams, and why measuring business outcomes, not AI usage, is the real indicator of success. He also shares how Celonis is helping organisations build the operational context required for AI to drive enterprise-wide transformation.
TAM: Celonis recently published a report stating that while 85% of business leaders want AI to run their business, 76% believe poor business processes are preventing them from doing so. Why does this gap exist? Is data readiness the primary challenge?
Manuel Haug: I think it is a combination of factors. Data readiness is certainly one part of the equation, but equally important is whether the business itself is ready for AI. When we talk about business processes, we use data to reconstruct what is happening inside an organisation, but ultimately it is about how a business actually operates. The biggest challenge we see is that business operations are often highly siloed. Teams work independently, technology stacks are fragmented, and supporting infrastructure is disconnected. When organisations try to introduce AI agents that need to operate across these silos, they struggle because the operational foundation is not designed for collaborative intelligence. AI can only scale when organisations themselves are ready to work in an integrated manner.
TAM: Many enterprises are planning to deploy multiple AI agents across different departments. However, departments themselves often work in silos. Doesn’t this increase the risk of AI making mistakes at scale?
Manuel Haug: Absolutely. One of the realities of AI is that if something doesn’t work well with humans, it is unlikely to work well with AI either. AI agents expose existing weaknesses within organisations. Human employees often compensate for fragmented processes by coordinating across teams and filling operational gaps. AI agents cannot do that automatically. Instead, they reveal where visibility is missing, where accountability is unclear and where teams lack alignment. In many ways, the challenges remain the same as they were years ago when organisations were improving business processes with human teams. AI simply amplifies both strengths and weaknesses. Without shared visibility and common operational goals, deploying more AI agents only magnifies existing problems.
TAM: AI is capable of much more than content generation or productivity tasks. How is Celonis helping organisations unlock AI’s full business potential?
Manuel Haug: The first step is creating an environment where humans and AI agents can work together effectively. We are moving towards a future where AI agents will manage significant parts of business operations, but that transition will take time. For the foreseeable future, organisations will operate with hybrid teams comprising both humans and AI agents. Our focus is on building the operational foundation that enables this collaboration.
Through our platform, multiple teams gain shared visibility into business operations. AI becomes another stakeholder within the business process, it participates in strategy reviews, operational monitoring and governance alongside people and enterprise systems. Unlike traditional automation, AI agents require continuous management. Organisations need to train them, improve them over time, document processes and treat them much like human teams rather than software that is deployed once and forgotten.
TAM: Traditional enterprise software is increasingly being challenged by AI agents. Do you believe conventional software will eventually become little more than data repositories?
Manuel Haug: I believe the reality is more nuanced. Enterprise software serves multiple purposes. It stores business data, preserves institutional knowledge and provides interfaces that employees are already familiar with. Some software platforms primarily function as systems of record with very limited business intelligence built into them. Those platforms may indeed face greater disruption. However, enterprise applications that encode deep operational knowledge actually have an opportunity to become even more valuable. That embedded expertise can be leveraged to make AI agents significantly more capable and effective.
TAM: If business context becomes the most valuable asset in enterprise AI, how does Celonis plan to maintain its leadership position?
Manuel Haug: Celonis occupies a unique position because we sit on top of operational process data, the information that explains how a company actually runs. Rather than focusing only on master data, we understand how decisions are made, how suppliers are selected, how customer service is delivered and which operational actions produce successful outcomes. On top of this operational context, we add business rules, governance policies and predictive capabilities. Since we understand how processes have performed historically, we can forecast future outcomes, anticipate delays and recommend better decisions. There will eventually be multiple context layers across the enterprise, but our strength lies in providing the operational intelligence that helps organisations understand and continuously improve how they function.
TAM: AI adoption requires significant investment. What advice would you give technology leaders to maximise return on investment?
Manuel Haug: Organisations should avoid measuring AI success through usage metrics such as the number of active agents or chatbot interactions. The real question should always be: what business outcomes did AI deliver? Leaders should evaluate AI by measuring improvements in automation, operational efficiency, customer satisfaction or cost reduction relative to the resources consumed. Whether it is better customer experiences, lower operational costs or improved business performance, AI investments should always be tied directly to measurable business value rather than technical adoption.
TAM: Is there anything we haven’t discussed that you believe is critical for organisations preparing for enterprise AI?
Manuel Haug: One aspect I believe deserves greater attention is the need for AI to contribute strategically, not just operationally. At Celonis, we focus on four critical capabilities: providing real-time visibility into operations, preserving organisational process memory, embedding business rules and governance, and enabling predictions and simulations for future decision-making. This allows AI agents to move beyond simply executing assigned tasks. They can evaluate different scenarios, recommend operational improvements and help organisations refine business policies based on data-driven insights. Ultimately, AI should become an active contributor to business strategy rather than just another automation tool. That is where we believe the future of enterprise AI is heading.
