Julian Theis is Industrial Co-Lead of the Object-Centric Event Data (OCED) Standardization Initiative within the IEEE Task Force on Process Mining and a Manager at the Deloitte Center for Process Bionics in Berlin. He shares with us his perspective on the future of OCED, his vision for the Working Group, and insights from his journey at the intersection of Process Mining and AI.

Julian, tell us a bit about yourself and how you entered the world of process mining.

My journey into process mining started during my PhD at the University of Illinois at Chicago in 2017, although I had already been in touch with the technology since 2015. In the US, I conducted research at the intersection of process mining and artificial intelligence. More specifically exploring how deep learning models can be enriched with process-aware structures to learn from process behavior more effectively. I worked on several applied research projects sponsored by industry partners, including one with one of the largest health insurance providers in the United States. That work made me realize two important things: first, how incredibly rich and underused enterprise process data truly is. And second, that meaningful AI in organizations requires a solid and expressive process representation.

After finishing my PhD, I moved into consulting to apply these ideas at scale for some of the largest organizations globally. Today, as a Manager at the Deloitte Center for Process Bionics, I lead teams developing smart applications that are running on top of real-time object-centric process mining architectures. Across industries, I’ve seen how object-centric perspectives unlock insights that would never emerge from classical, case-centric logs. That experience is ultimately what brought me to the OCED standardization initiative.

What sparked your interest in object-centric process mining and OCED?

In many real-world implementation projects, you quickly hit a ceiling with traditional event logs. Enterprise processes don’t revolve around one case notion. They revolve around interconnected objects like orders, items, deliveries, inventory batches, equipment, master data, and more. Over the last few years, I focused especially on our automotive and manufacturing clients, developing solutions for their core production and supply chain processes. The variety of possible case notions is almost limitless. 

To truly reflect the complexity and to optimize and steer processes, we often had to engineer custom multi-object models, because no standard format could capture those relationships properly. The moment you try to connect P2P, O2C, and inventory flows across multiple systems, that limitation becomes glaringly obvious. Object-centric data is simply a more truthful representation of how businesses operate. OCED is the missing foundation for scalable, interoperable object-centric analyses and the foundational layer for something I like to call Enterprise OS. By Enterprise OS, I mean a unified layer that standardizes process data across all systems and enables AI, automation, and all other kind of intelligent applications to operate reliably on top of it. And that’s why I’m passionate about it.

As the industrial co-lead of the OCED Working Group, what do you see as your mission?

My mission is twofold and deeply informed by my background in both research and industry. On one hand, I bring the industry perspective. Drawing from years of experience designing and deploying object-centric process mining architectures, I help ensure that OCED addresses the real challenges faced by practitioners, vendors, and end users. This includes considerations around scalability, integration across heterogeneous systems, and practical deployment constraints. Those are issues that are particularly prominent with the clients I work with.

On the other hand, I serve as a facilitator and translator, bridging the gap between academic rigor and industrial pragmatism. My academic background provides the needed foundation in structured, methodical thinking, enabling me to critically assess proposals and ensure the standard is logically sound. At the same time, my consulting work often involves complex, multi-stakeholder environments where priorities and perspectives differ. This combination allows me to guide discussions, reconcile diverse viewpoints, and ensure that decisions are both practically implementable and aligned with real-world business needs. 

A good standard must be rigorous enough to be reliable, yet pragmatic enough to be adopted. My role is to help the group strike that balance. By guiding discussions, surfacing conflicting views early, and ensuring the Working Group converges toward a shared, usable outcome.

What is your approach to leading the Industrial side of the OCED Working Group, and how will this influence how the group operates?

My approach is centered on combining structure with practicality. On the one hand, I believe a working group functions best when there is clarity. Clear goals, clear meeting rhythms, and transparent decision-making. On the other hand, OCED is deeply connected to real-world operational challenges, so we must constantly validate our ideas against practical use cases and implementation experience.  I also place strong emphasis on creating a standard that people can understand. Technical correctness is not enough. Adoption depends on intuitive concepts, consistent terminology, and examples that show how OCED works in actual processes. Finally, I see sustainability as essential. OCED should evolve with the community, not be a static release. My goal is to help build a collaborative environment where contributions from practitioners, vendors, and researchers continuously move the standard forward.

Why, of all topics, focus on data standardization?

This is a question I hear a lot. Let us consider a glamorous field of this time, AI. Artificial Intelligence is exciting, but the biggest return on investment for AI in organizations doesn’t come from the AI model itself. It comes from the structure and quality of the underlying data, the context you provide to the AI. Without a clear, consistent, and scalable representation of the processes where AI will ultimately be deployed, even the most advanced models can only scratch the surface.

OCED provides the layer that makes real AI possible inside enterprises. It builds the foundation for cross-object predictions, process-aware machine learning, simulation, digital twins, or advanced decision-support systems. More broadly, OCED serves as a foundational layer for AI. Meaning, OCED is not an alternative to AI. It is a prerequisite. Organizations that want to leverage AI across end-to-end processes need an extensible, standardized way to represent object relationships. That is where real value lies. And where organizations ultimately achieve the highest return on investment.

What most excites you about the future of process mining as OCED matures?

I’m excited about the shift from isolated process analyses and applications toward integrated process ecosystems, where data flows consistently across domains. With OCED, companies can connect processes across systems using standardized, object-centric representations, without reinventing the wheel each time, allowing tool vendors and practitioners to innovate on a shared foundation.

In the long term, I see process mining evolving into the Enterprise OS, which will be a unified framework combining process design, execution, monitoring, compliance, and AI. OCED acts as the kernel of this Enterprise OS, providing the standardized process language that enables scalable, interoperable, and intelligent process management across the enterprise.

What challenges do you foresee for OCED adoption?

The main challenge is diversity. Multiple industries, vendors, and solution providers all have different expectations. Achieving consensus requires disciplined collaboration. Another challenge is bridging legacy landscapes. Many companies operate on complex, heterogeneous architectures. Providing good guidance, templates, and reference models will be critical. And adoption requires trust. The community must feel that OCED is inclusive, transparent, and not driven by any single stakeholder group. There is a lot of work to be done.

What are the next steps for the Working Group?

The Working Group is currently focused on developing the personas that the standard needs to address and on gathering domain use cases. This work helps us define the scope of OCED in a way that is practical and relevant for a variety of stakeholders. Importantly, this is not done in isolation. We actively involve the Extended Working Group, which is being set up, as well as the broader community through our regular community calls.

Once the scope and use cases are well-defined, the next major milestone will be opening a stable draft to the broader community. However, there is still substantial work to be done to reach this stage. We have divided the standard into different dimensions, each of which needs to be elaborated against the personas and use cases to ensure the draft is both robust and has a high likelihood of adoption. The road ahead is still long, but every step brings us closer to a standard that makes a lasting impact on the way organizations understand and optimize their processes.

What motivates you personally to invest in OCED?

For me, the motivation is simple. I want to help create the foundation that allows process mining and AI to scale inside organizations. Over the years, I’ve seen process mining teams struggle with fragmented data models, incompatible formats, and duplicated engineering effort. A shared, object-centric standard can remove those barriers, so we can focus on what truly matters: Understanding, improving, and innovating processes.
Being part of shaping such a foundation feels both meaningful and impactful.