In the video below the post, Sharam Dadashnia walks through the main highlights of Scheer PAS 26.0 up to 26.0.2. Below is the written companion, with a bit more context so you can relate the changes to your own platform setup and upgrade plans.

More control and visibility for AI agents in Kubernetes operations

Scheer PAS 26.0 strengthens operational governance for AI agents by extending PAS Administration with a dedicated AI Agent Configuration Panel. The idea is simple: agent workloads should be managed with the same level of clarity as xUML services. In the new panel, administrators can centrally view and edit agent configurations, and they can immediately see in which service an agent has been created. This reduces the typical “where is this agent coming from” back and forth that often happens once multiple teams start deploying agents in parallel.

To keep the workflow consistent, AI agents are also integrated into the xUML Service Details. A new AI Agents section lists the agents created within a service, including name and description, and it offers direct links into the configuration view. For operations teams, this means less searching and fewer context switches when validating or adjusting workloads.

Trace-logging as a foundation for trust

Alongside the UI enhancements, PAS 26.0 introduces logging of agent traces. This is an important step for anyone who needs end to end traceability of agent executions, whether for troubleshooting, auditability, or simply for understanding why a certain outcome occurred. As agent usage grows, traceability becomes one of the key building blocks for keeping Agentic Process Orchestration manageable in production.

Runtime-switching made easier for Docker and Kubernetes installations

Keeping services up to date often requires careful sequencing. If multiple PAS Runtime versions are available in your installation, the new Update Runtime feature allows you to change the runtime version used by a specific xUML service through the Administration UI. This supports controlled, targeted updates without turning every runtime change into a broad platform event.

Designer productivity: fewer clicks for adapter based integrations

Beyond operations, there are also noticeable improvements for everyday modeling work. Starting with PAS 26.0, when you drag and drop adapter operations such as REST, OData, or SAP RFC into an activity diagram, the required extension is applied automatically. It may sound like a small change, but it adds up to significant time savings, especially in projects with many integration steps and recurring modeling patterns.

In addition, this version brings further enhancements in the Designer, including SAP RFC adapter support for dynamic connection parameters, for example to use different SAP users per call, as well as the option to selectively exclude methods during xUML Java import. Updated examples, including extended OData usage and new scenarios for synchronous and asynchronous service communication, round out the package.

Persistent state: migration notes and fixes

Because persistent state is a critical runtime topic, there are two migration notes to keep in mind. With PAS 26.0, the persistent state schema version is updated to version 17 (18.1 for MariaDB). After upgrading, downgrading to an older runtime is only possible if the persistent state database is empty. With PAS 26.0.1, the ability to use a separate database for Persistent State external attributes has been removed, linked to a fix in how existing persistent state objects are updated. PAS 26.0.2 then delivers an additional fix in the xUML Runtime regarding persistent state handling.

For the full list of changes across Administration, API Management, Analyzer, community assets, Bridge runtime updates, and more, please check the complete release notes linked alongside this post and the video.

Video - German

Video – English

Scheer PAS 26.0 – Release Notes

The post Echter Mehrwert durch KI-Agenten: Scheer PAS 26. Produktupdate appeared first on Scheer PAS.

Why traditional process automation reaches its limits

Die meisten Unternehmen betreiben ihre Kernprozesse entweder auf klassischen Automatisierungsplattformen oder innerhalb von Standard-Business-Applikationen wie ERP-Systemen. Diese Umgebungen wurden vielfach modernisiert, doch das Grundmuster ist häufig gleich geblieben: Fachbereiche definieren Prozesse konzeptionell, IT setzt sie technisch um, vorwiegend als deterministische, regelbasierte Modelle. In der Praxis erzeugt diese Trennung Reibung. Abstimmungszyklen zwischen Fachbereichen und IT führen zu Verzögerungen und Missverständnissen, während fest verdrahtete Logik Prozessänderungen zu Projekten macht statt zu kontinuierlicher Verbesserung.

Das Ergebnis ist bekannt: Prozesse funktionieren gut, bis sich der Kontext ändert. Wenn neue Anforderungen entstehen, kann das System nicht in Echtzeit reagieren. Organisationen kompensieren das mit manuellen Workarounds, eskalierenden Change-Requests und längeren Durchlaufzeiten.

What agentic orchestration changes

Scheer PAS is taking a next-generation approach and enables process solutions based on AI agents.These agents use large language models to reason, plan, and execute multi-step workflows in a goal-directed manner, adapting their actions to changing conditions. Instead of encoding every decision in static rules, companies can work with goal-based execution that supports continuous optimization over time.

The platform also supports collaboration among multiple AI agents and provides a global orchestration layer for end-to-end business processes. This multi-agent coordination is intended to keep processes resilient even when exceptions occur, without sacrificing speed and transparency in operations.

Built for integration, governance, and day-to-day operations

Agentic orchestration only creates value if it can be integrated into existing system landscapes. Scheer PAS addresses this through deep API connectivity to established SaaS solutions and enterprise applications, enabling agents to access the data and process capabilities already embedded across the application landscape. Standard applications thus become reliable data and process providers, while a new orchestration layer brings deterministic and probabilistic behaviors together in a coherent flow.

The platform supports AI standards such as MCP and A2A und umfasst Funktionen für Echtzeit-Monitoring, Protokollierung und Visualisierung von Aktivitäten, Agentenverwaltung sowie einen integrierten Coding-Agent zur Unterstützung bei der Codegenerierung.

As Prof. Dr. August-Wilhelm Scheer summarizes, the future will not be shaped by static workflows alone, but by intelligent systems that reason, adapt, and collaborate.

All details about the announcement can be found in the full press release on our partner’s website: [Link to the full press release].

The post Über regelbasierte Automatisierung hinaus: Agentic Process Orchestration appeared first on Scheer PAS.

The Growing Digitalisation Backlog

Many organisations now have an extensive backlog of initiatives:

• Process automation
• AI initiatives
• Data platforms
• Integration projects
• Internal applications for business departments
• Modernisation of existing systems

Every department recognises the potential of digital technologies to accelerate processes or unlock new opportunities. The problem is not a lack of vision. The problem is implementation capacity..

Digitalisation teams today are under significant pressure. At the same time, they must:

• Operate existing systems
• Evaluate new technologies
• Enable AI initiatives
• Ensure governance and compliance
• Build trust across the organisation
• Implement projects requested by business departments

Newly established digitalisation teams in particular operate in a challenging environment. They are expected to deliver results while simultaneously building the organisational foundation for digital transformation.

The result is a structural bottleneck. The backlog continues to grow, while expectations for faster implementation keep increasing.

The New Reality of Software Development

At the same time, a technological development is fundamentally changing the playing field. Large language models specifically trained for code generation are evolving rapidly.

Today they are capable of:

• Generating complete services
• Creating integrations
• Developing user interfaces
• Writing test cases
• Supporting deployment pipelines

With the same resources, organisations can now implement significantly more solutions than just a few years ago. The barrier to entry for software development is decreasing. But this is exactly where a dangerous misconception emerges. Because Software zu bauen, ist nicht dasselbe wie Enterprise-Software nachhaltig zu betreiben.

Sustainable Software Is Strategic Infrastructure

Anyone who has operated an application in production for several years knows: The first release is only the beginning.

Enterprise software must continuously:

• Be extended
• Be maintained
• Be connected to new systems
• Meet compliance requirements
• Run reliably
• Scale

A prototype can solve a problem quickly. But without a solid architecture, it quickly becomes a long-term burden. In many organisations, this creates a new problem: Maintenance effort eventually exceeds the original development effort.

Software Investments Are Strategic Investments

Companies often still treat software projects like traditional IT projects. But software has long since become strategic infrastructure. The comparison with industry is fitting.

When a manufacturing company purchases a new production machine, several key questions are asked:

• Can the machine increase production output?
• Does it improve quality?
• Can it open up new customers or markets?
• Will the investment pay off in the long term?

The same questions must also be asked when investing in software. The real ROI is not created by prototypes. It is created by stable, productive systemsthat continuously deliver value.

The Missing Platform Layer

Many organisations are currently trying to address their backlog solely by accelerating software development. However, speed alone is not enough. What is missing is a platform that orchestrates the entire digital transformation.

Such a platform must cover several dimensions:

• AI-assisted development
• Integration of existing systems
• Process automation
• Orchestration of AI agents
• Governance and compliance
• Operations and monitoring

Only when these elements work together can organisations sustainably implement their digital initiatives.

Time-to-Value Becomes the Key Metric

The central question is therefore no longer: “Can we build this solution?” With modern AI tools, the answer is almost always: yes. The crucial question is: How quickly can we turn an idea into real business value?

In many organisations, the journey from
Idea → Prototype → Productive Software → Business Impact
still takes months or even years.

In a world of exponential technological development, that is simply too slow. Companies that reduce their time-to-value gain a significant competitive advantage.

Agentic Process Orchestration as an Approach

Agentic Process Orchestration combines several crucial elements:

• AI-assisted development
• Enterprise Integration
• Process automation
• Orchestration of intelligent agents

Instead of building isolated solutions, this approach creates a platform that enables organisations to systematically implement digital initiatives.

How Scheer PAS Supports Companies

At Scheer PAS, we support organisations exactly in this scenario. Companies face the challenge of implementing a large number of automation and digitalisation initiatives.

Our platform enables organisations to:

• Develop solutions faster
• Seamlessly integrate existing systems
• Orchestrate processes across system boundaries
• Deploy AI-driven automation productively

The key effect is clearly measurable:

Time-to-value decreases dramatically.

Conclusion

Digital transformation is no longer a question of ideas. There are plenty of ideas. The decisive factor is execution speed. Companies that can turn their backlogs into productive solutions will be the winners of the next decade. Digital leadership today therefore means one thing above all:

Consistently reducing time-to-value.

More information about using Scheer PAS AI Agents can be found here.

The post Warum „Time-to-Value“ zur entscheidenden Kennzahl der digitalen Transformation wird appeared first on Scheer PAS.

This AI-based showcase demonstrates how implicit expert knowledge can be systematically digitized and transformed into a scalable corporate memory. As a consistently process-oriented digitization platform, Scheer PAS provides the backbone of the solution: by seamlessly combining integration, automation, orchestration and agent-based AI, complex planning and decision-making processes become transparent, reproducible and easier to scale. With the flexible process control of Scheer PAS, the balance between rule-based workflows and AI agents can be precisely adjusted – building trust in AI-supported processes and enabling future-proof, knowledge-based production.

Read more: Scheer PAS and German Edge Cloud present AI showcase

The post Scheer PAS und German Edge Cloud präsentieren KI-Showcase auf der HANNOVER MESSE: „Digital Industrial Engineer“ appeared first on Scheer PAS.

Christian, it’s great to have you back for another conversation. There are several topics we want to touch on today, but we’d like to focus especially on user interaction, interfaces, and UI in the context of process automation. From your perspective, what should organizations be preparing for as we move toward 2026?

Christian Linn

Gerne! Ich denke auch, es gibt einige neue Trends zu besprechen.

Sebastian Dietrich

When we talk about process automation, we often have to revisit terminology that has been used for years. One example is the concept of the citizen developer. How relevant is it today from your point of view? Is it still something you encounter frequently in the market, or has its importance faded?

Christian Linn

The citizen developer is a concept that has evolved over the past few years and across different domains. At its core, it’s about enabling people without formal programming or software engineering training to develop applications. In principle, that idea has merit, but not across all use cases or application areas.

As with many topics in the IT and technology space, expectations were very high when the concept first gained attention. Over time, however, those expectations have only been partially met. The reason is relatively straightforward. There are scenarios where applications created through Low-Code approaches by citizen developers can work well and deliver value. However, once we move into business-critical, complex processes and systems, these approaches quickly reach their limits.

What ultimately matters is not how fast an application can be clicked together, but whether it is based on a scalable, secure, and maintainable architecture, and whether there is a sound software concept behind it. This is where citizen developer approaches often fall short.

Sebastian Dietrich

Another concept that appears frequently, especially in external communication when tools are positioned as beginner-friendly, is graphical drag-and-drop interfaces. On the other end of the spectrum, we now see the emergence of vibe coding, which works with natural language and AI. Are these two extremes, or do they simply address different user groups?

Christian Linn

They primarily address different user groups, even though the underlying goal is quite similar. Both approaches aim to significantly reduce the time required to develop software applications.

Low-Code approaches with graphical user interfaces tend to appeal more to citizen developers, but they can also support professional developers in certain scenarios by enabling faster implementation. Vibe coding, on the other hand, is a relatively new approach where software development is guided by AI through natural language or voice input.

Both approaches come with advantages and limitations. In practice, especially in the context of process automation and enterprise software, the reality will likely settle somewhere in between.

Sebastian Dietrich

If these approaches shorten the path from idea to application for different target groups, that’s clearly attractive. At the same time, applications and architectures still need to be validated and optimized. Does this mean we’ll see a resurgence of Pro-Code development, or does it simply highlight that without deep coding expertise, we can’t fully understand the bigger picture?

Christian Linn

That question is closely related to whether we will still need software developers who are able to write code. My answer is very clear: yes, absolutely.

Vibe coding tools and coding assistants primarily help accelerate code writing and application development. Where they still struggle, and where deep expertise remains essential, is in architecture and technical design. This includes modularization, complex system design, and understanding how an application behaves as part of a larger ecosystem.

Writing code is only one aspect of software development. These tools can support that part very effectively. However, when it comes to business-critical software, you need the experience to design robust architectures, ensure security, performance, and long-term maintainability. That expertise cannot be replaced.

What we will likely see is that AI-supported coding starts to challenge traditional Low-Code approaches. Simple applications, demos, proofs of concept, and prototypes can already be created extremely quickly through natural language interfaces. In many cases, this is even faster than what classic Low-Code tools can offer. I do expect a noticeable shift in that direction.

Sebastian Dietrich

So looking ahead to 2026, would you say we’re moving toward a sweet spot between Low-Code and professional, Pro-Code frameworks?

Christian Linn

I think it’s helpful to differentiate between three distinct areas. Traditional Low-Code approaches are well suited for applications with limited criticality, especially when there is sufficient governance and compliance in place to operate them safely within a corporate environment. In those cases, they can be a good way to implement departmental use cases.

Vibe coding, on the other hand, is particularly strong in demos, proofs of concept, and prototypes. It enables teams to validate ideas quickly, gather feedback, and iterate. That makes it very effective for early-stage exploration.

When we talk about productive enterprise systems, including central IT systems and core business processes, we still need a combination of AI-supported code generation and hands-on Pro-Code development. The expertise of experienced software developers remains essential to ensure that these systems meet the requirements of enterprise IT infrastructures.

Sebastian Dietrich

If we look across all levels, from platform vendors to end users building small applications, we know that development itself is only part of the work. A large share of the effort goes into debugging and testing. Is this an area where AI agents are already delivering clear added value?

Christian Linn

Absolutely. AI already supports many different phases of the software lifecycle. We’ve talked a lot about development, such as code generation, test generation, documentation, and translating requirements into technical specifications.

Beyond that, AI and AI agents are increasingly supporting the operational side of software. This includes monitoring, error analysis, and in some cases even autonomous debugging. In certain scenarios, AI can identify issues, correct them, and redeploy updated versions of software automatically.

There are many areas where AI agents are already providing real value in software operations. This is also a direction in which we plan to further expand the Scheer PAS platform in 2026..

Sebastian Dietrich

So in short, agentic AI brings more control and support, rather than simply producing outputs without oversight.

Christian Linn

Exactly. In enterprise software, development is only one part of the equation. Operation and maintenance are just as critical. If AI can support both areas while still allowing organizations to retain control, then it becomes truly valuable.

Sebastian Dietrich

That’s a great way to wrap it up. Thank you, Christian.

The post Die Zukunft der IDE in der Prozessautomation – Ein Interview mit Dr. Christian Linn appeared first on Scheer PAS.

Hi Sharam, good to have you! Today we’ll be talking about some interesting developments in the world of RPA and, of course, AI, which will shock nobody. The first question I’d like to ask you is: In your opinion, how has Agentic AI changed, or perhaps even replaced, classic robotic process automation (RPA) in recent years, and how do you see this evolving in 2026?

Sharam Dadashnia 

That’s a very interesting – and in 2026, entirely expected – question, and there are two key aspects to consider. The first is integration. Traditional RPA relied on software robots interacting with user interfaces, essentially mimicking human behaviour across existing systems. In a nutshell, bots automated processes by clicking and typing instead of humans. For many years, this served as a bridging technology.

“Classic RPA was a bridging technology — Agentic AI addresses the reality of knowledge work and decision-making.” 

However, modern software architectures now provide APIs directly, making this UI-based automation largely obsolete. This shift is reflected in the market; the major RPA vendors have seen declining valuations and revenues over the past several years. Classic RPA is not only increasingly unnecessary but also inherently rigid. Rule-based automations struggle to adapt to changing user interfaces or dynamic environments.

The second aspect relates to knowledge work. In real business processes, decisions are rarely simple rule executions. Humans gather information from multiple sources, assess context, make judgements, and then communicate outcomes back to systems. This is precisely where Agentic AI adds significant value.

Agentic AI can support, or even fully automate the preparatory phase of decision-making by aggregating data, analysing documents, and presenting context-aware recommendations. Many companies currently favour a human-in-the-loop approach, where the AI prepares decisions and humans retain final responsibility. This is understandable, given that no business is ready to delegate accountability entirely to AI.

To give a concrete example from the insurance sector: case processing that previously took 20 to 30 minutes can be reduced to around five minutes when all relevant information is analysed and prepared in advance. Once organisations gain sufficient trust in these systems, full automation becomes a realistic next step.

Sebastian Dietrich 

Completely understandable. However, you’ve mentioned trust several times, which is especially relevant for regulated industries. Many of these solutions rely on cloud technologies. With everything going on in the world between the EU, the East, and the West, how do you see trust evolving in this context, for example, with initiatives like the EU Sovereign Cloud?

Sharam Dadashnia 

In practice, we see two dominant patterns. On the one hand, many AI services are still US-based. On the other hand, US hyperscalers now operate data centres within Europe, including Germany, enabling highly secure, isolated AI solutions that comply with EU sovereignty requirements.

The key point is that AI adoption is always use-case driven. Not every scenario requires maximum computational power from a service located outside the EU. In many cases, we have already implemented dedicated, protected AI systems, particularly for the public sector and other highly regulated environments, that run entirely within controlled data centres and are fully GDPR-compliant.

When working with public or non-sensitive data, the regulatory hurdles are naturally lower. Each project requires an individual assessment of data sensitivity, compliance needs, and technical requirements. Importantly, organisations today have real choice: European providers like Mistral, as well as open-source models, can be integrated and operated as managed services. We deliberately avoid exclusivity with any single provider.

Sebastian Dietrich 

That’s good to hear since providing choice in times like this is crucial for a lot of companies hoping to implement new technologies. Let’s talk about AI models and regulations: With the EU AI Act setting new guidelines, do you see this as limiting innovation and automation, or as a necessary corrective step?

Sharam Dadashnia 

This seems like a sensitive subject, but I don’t think it should be. From my perspective, the AI Act, coming fully into force this year, introduces a necessary framework. It forces organisations to actively assess whether an AI use case is highly critical, sensitive, or relatively low-risk.

“The EU AI Act doesn’t limit innovation; it creates clarity and trust.”   

I see this as a positive development amidst this huge AI hype. It helps prevent careless handling of sensitive data, such as feeding confidential information into unsecured systems. The regulation doesn’t limit innovation; it adds transparency. Essentially, it acts as a labelling mechanism that helps organisations understand which services are appropriate for which use cases.

With this clarity, it becomes easier to design compliant solutions, something we are already doing successfully with our customers today. We don’t see the AI Act as an obstacle, but as an additional layer of assurance.

Sebastian Dietrich 

Good to hear that, because it seems that a lot of the AI providers tend to look differently on EU AI Act. As a platform provider, how do we ensure technological sovereignty over business processes while integrating existing solutions and remaining future-proof?

Sharam Dadashnia 

Our platform allows us to maintain full transparency and control at every step of a process. We can see exactly where data originates, how it is processed, and where it is sent.

If an organisation is using AI services outside the EU, such as certain US-based APIs which everyone already knows, we can identify this and make it visible. This enables active compliance monitoring and informed decision-making. In some cases, it may be perfectly acceptable; in others, it may require reassessment from a regulatory or risk perspective.

This transparency is crucial. It allows organisations to understand their current setup and make deliberate choices about sovereignty, compliance, and future architecture.

Sebastian Dietrich 

That’s a nice way to put it but there is a growing need for even more technological advancement. We see growing interest in multi-agent systems and open-source approaches. Are companies actively waiting for a European open-source AI movement to mature?

Sharam Dadashnia 

Many organisations are already experimenting with AI agents in isolated use cases. Broadly speaking, we have seen two approaches.

The first is a disruptive one, where companies fundamentally rethink entire divisions or processes with an AI-first mindset. That seems like a very bold approach. The second, far more common in German-speaking countries, is an evolutionary approach, where existing processes are gradually enhanced with AI.

While the strategic intent may already exist, large-scale implementation is still rare. Most organisations start small to build confidence and trust. Our approach is to identify process steps where the balance between effort and impact is optimal. These projects deliver quick wins, build trust in the technology, and create momentum for larger initiatives.

Start-ups often adopt AI-first strategies from day one. Established enterprises tend to move more cautiously, but steadily.

Sebastian Dietrich 

There also seems to be a preference for starting with a proven vendor rather than going directly into open source.

Sharam Dadashnia 

That largely depends on organisational maturity and internal capabilities. Building in-house expertise takes time, especially in a relatively new discipline where experienced talent is scarce.

Large enterprises often explore both paths, open source and enterprise solutions, in parallel. We ourselves leverage open-source models where appropriate, but we also recognise the complexity involved. Having worked with AI-driven process automation for over five years, we have already overcome many of the early challenges.

Choosing an experienced vendor can significantly reduce time, cost, and risk. Just as with any complex system, professional support becomes indispensable once scale, reliability, and compliance are required.

Sebastian Dietrich 

Finally, do you have any closing thoughts on 2026 for the topics we’ve discussed today?

Sharam Dadashnia 

For me, 2026 is a decisive year for AI adoption in German-speaking countries and across Europe. While public discourse often paints an overly optimistic picture, the reality is that many organisations are still in early stages.

At the same time, workloads are increasing and capacity is limited. Demographic change, especially the retirement of the baby-boomer generation, will lead to a significant loss of expertise. AI and process automation are essential tools to close this gap without compromising quality.

“Europe should not become a digital colony of the US or Asia. That is the mission we are pursuing, and the time to start is now.”

If Europe fails to act now, others will move faster. My goal is clear: we must remain competitive, independent, and confident in our own solutions. Europe should not become a digital colony of the US or Asia. That is the mission we are pursuing, and the time to start is now.

The post Experteninterview: Warum Europa keine digitale Kolonie werden darf appeared first on Scheer PAS.

1. Agentic AI 2025: From Experiment to Platform Capability

Question: Looking back at 2025: Which developments in the field of Agentic AI for process automation were the most impactful for you as Head of Development?

Dr. Christian Linn:
2025 was exciting for us in two dimensions: from our customers’ perspective and from a technological perspective.

On the product side, we introduced a new feature to the Scheer PAS platform that enables our customers to build their own AI agents and integrate them into their process automation—either for complete processes or selectively for sub-processes. This was an important step: moving away from pure assistant scenarios toward agents that actually take over tasks and assume execution responsibility. This gives our customers significantly more dynamic and flexible automations.

At the same time, there was a lot of movement on the technological side. Of course, new and more powerful models from the major providers emerged. But more decisive for us was the shift in focus within the Agentic AI space: away from “let’s quickly build an agent” toward “how do we make agents truly production-ready.” Topics such as security, traceability, and transparency of execution have moved much more into the foreground. These are exactly the aspects required to deploy agents in more critical process environments—and aspects we have considered in Scheer PAS from the very beginning.

2. “Nice Demo” or Real Scalability?

Question: Many companies experimented with AI agents in 2025. From your perspective, what made the difference between “nice demos” and truly scalable process automation?

Dr. Christian Linn:
We saw many projects where an initial agent prototype was quite impressive—but failed to make the leap into continuous productive operation. Based on our experience in process automation, there are a few clear reasons for this.

A key success factor is not viewing the agent’s use case in isolation.A key success factor is not viewing the agent’s use case. An agent never operates in a vacuum. It is always part of a business process. That means: What happens before the agent? Which systems provide context? What happens afterward, once the agent has completed its task? Who intervenes if something goes wrong?

This brings two points into play:

1. Deep system integration:
   Agents must be able to integrate in a controlled manner into existing IT systems. Only if they can access relevant data and functions—via stable APIs and well-defined interfaces—does real value emerge beyond the demo stage.
2. Robust orchestration and governance:
   We see BPMN-based orchestration as an important anchor of stability. Processes remain transparently modeled, even when agents make dynamic decisions. In addition, clear governance rules are required: Which actions is an agent allowed to execute? Where are approvals required? Which data is the agent allowed to see?

And one more thing: companies need realistic expectations of LLMs such as Mistral. AI agents are not a silver bullet. There are scenarios where they are very helpful—and others where classic, rule-based automation is perfectly sufficient. The projects that successfully went into productive operation in 2025 found this balance.

3. Orchestrating Multiple Agents: The Critical Challenges of 2025

Question: From a product and architecture perspective: Which challenges in orchestrating multiple AI agents were most critical for you to solve in 2025?

Dr. Christian Linn:
We clearly assume that in many scenarios, multiple agents will work together. The key question is: How do you orchestrate this interaction reliably?

Currently, we see two main approaches:

• Multiple agents are embedded into a rule-based process structure , for example BPMN processes. The sequence, handover points, and escalation paths are clearly modeled.
• A multi-agent framework handles the coordination itself: a higher-level agent or an LLM controls which agent takes over which part of a task.

Both approaches have potential but place high demands on product and architecture. In 2025, the following challenges were particularly critical for us:

• Coordination and context propagation:
  Agents must know which step of the overall process they are in and which context they are building upon. Loss of context quickly leads to wrong decisions or unnecessary loops.
• Error handling and robustness:
  What happens if an agent cannot execute a task, a system is unavailable, or unexpected data is received? Robust error paths, automatic fallbacks, and test scenarios are essential here.
• Security and permissions:
  Classic enterprise topics such as authorization management do not become less important with agents—quite the opposite. We must clearly define which agent has access to which data, systems, and actions.

At Scheer PAS, we worked intensively throughout 2025 to map these aspects into the platform: from model-based orchestration to monitoring and logging, all the way to role and permission models for agents. Only in this way can agents truly be deployed at scale in large, critical processes.

4. From Pilots to End-to-End: Priorities for 2026

Question: Looking ahead to 2026: In your view, what will be the most important priorities for users who want to move from isolated AI pilots to end-to-end process automation with Agentic AI?

Dr. Christian Linn:
For 2026, I see four central priorities:

1. Selecting suitable use cases:
   Not every process is immediately a candidate for Agentic AI. Companies should start with clearly defined, business-critical yet manageable processes where the added value of AI is obvious—such as high manual effort or many exceptions.
2. Standardizing data and APIs:
   Without clean data models and well-documented APIs, any agent orchestration quickly reaches its limits. Anyone who wants to scale in 2026 must lay the foundation: integration, API management, and consistent interfaces.
3. Defining human-in-the-loop patterns:
   Humans remain an integral part of automated processes. Companies should consciously define when an agent makes decisions autonomously and when it provides recommendations that must be approved by business users.
4. Clear KPIs for value measurement:
   To move from pilot projects to broad adoption, solid metrics are required: cycle times, error rates, manual effort, employee and customer satisfaction. These KPIs should be considered from the very beginning.

5. Practical Advice for the Agentic AI Roadmap 2026

Question: If you had to give companies one piece of practical advice for their Agentic AI roadmap for 2026, what would it be?

Dr. Christian Linn:
My advice would be: Start focused—but start.

Choose a clearly defined, professionally relevant process where real added value becomes visible. Rely on an orchestration platform like Scheer PAS that supports you in areas such as integration, security, scalability, and monitoring. Combine this platform with a well-governed LLM such as Mistral that aligns with your compliance requirements.

And then: learn from real usage. Gather experience in development, operations, and the interaction between business and IT. These experiences cannot be replaced by any upfront analysis. Those who are ready in 2026 to start simply—with a solid foundation and a clear use case—will be able to integrate Agentic AI step by step into their end-to-end processes..

The Process iPaaS

Where Process meets Integration

Model Your Processes Rapidly!

The post Agentic AI in der Prozessautomatisierung: Rückblick 2025 und Ausblick auf 2026 appeared first on Scheer PAS.

However, the rising divide between the far west and far east in the way the AI is being developed and implemented left a lot of European companies split in the decision of which way is the right way to go.

Instead of choosing sides and opening their IT ecosystems to either side, there is a new player in the middle of the AI race which keeps the established European principles and abides to the introduced regulations yet accelerates the growth and development.

The Need for a Middle-Ground Solution

Ever since April of 2021, European Commission has been trying to get ahead of the rising popularity of AI technologies which saturated the markets without any proper control.

For enterprises which struggled (and are perhaps still struggling) to find a proper place for AI in their business processes, this only meant stricter regulations and slower development. Luckily, the effort in maintaining strict and secure regulations with the technology advancement have been made and ever since April of 2023, a solution for companies operating inside of EU has been formed.

Based in Paris, Mistral AI proved to be an excellent choice for companies that adhere to EU regulations the AI solution itself completely complies with EU AI Act meeting the security and privacy requirements.

With strategic partnerships with seasoned players in EU such as SAP, choosing Mistral AI as an LLM provider in an iPaaS or process automation platform already comes with some advantages for any European company.

The Matter of Choice

Die europäische Grundidee der Wahlfreiheit prägt auch unsere AI Agents in Scheer PAS. Wir schreiben keine einzelne Lösung vor, sondern bieten offene Unterstützung für verschiedene Modelle. Unternehmen, die Scheer PAS für Integration und Prozessautomatisierung einsetzen, können zum Beispiel Mistral AI als LLM nutzen – und so Agentic AI mit hoher Transparenz und Sicherheit im EU-Rechtsraum realisieren.

With its “Privacy-first” motto, Mistral AI supports the companies with their process automation and mining through Scheer PAS platforms unique offer including the complete vertical and horizontal IT system integration, API management, Low-Code approach and many others.

Of course, using AI Agents in PAS is not limited to Mistral AI as the LLM provider, which enables users to venture outside the scope of a single provider. Still, using AI Agents to further improve efficiency in business processes with complete adherence to standards and regulations provides a piece of mind for the users who value privacy and security.

European Solution for Competitive Enterprise

Besides choosing an LLM provider such as Mistral for AI Agents with security and compliance in mind, selection of a digitalization platform should also be considered with the current situation in the world.

Therefore, having a platform hosted on European servers, built and maintained in Germany and Switzerland, using an AI solution completely adhering to domestic principles and standards ensures enterprises with full transparency, ease of use and streamlined migration.

With all the capabilities of Scheer PAS, integration of IT landscape with process automation is done without losing sight on what matters: actual business processes and their efficiency. Try out the Agentic Process iPaaS platform today for free and leverage the AI agents in Europe without dealing with EU laws and regulations since they are all taken care of.

How Agentic AI Works in PAS

Since enterprise-grade process automation cannot rely solely on chat bots, Scheer PAS implemented the AI Agents as proactive part of the platform not only as a suggestive help in manual process setup and automation but as vital parts dealing with process exceptions and redundancies in business.

Therefore, since AI Agents in Scheer PAS actually deal with business processes and real flow of data, the LLM selection was of vital importance for companies inside of EU. Mistral AI as an LLM provider can then guarantee the security and privacy of information handled by the agents while Scheer PAS handles business processes and system integrations.

This combination allows users to not only handle the structured inputs from various systems integrated within the platform, but also to handle unstructured data and exceptions with complete transparency. If an exception occurs or redundancy is recognized by an agent, it is clearly communicated and before any action is taken, the relevant persons are informed with complete reports and clear suggestions.

This is considered to be the next step in process automation and system integration since collaborating agents take the proactive role of handling business processes which are not only deterministic and preset but agile and evolving.

Using Mistral AI with Scheer PAS provides a secure and private canvas for future development and agile of enterprises looking to leverage both AI agents and modern technology in order to improve efficiency and agility.

The post Sicher zukunftsfähig: Agentic AI in Europa mit Mistral AI & Scheer PAS appeared first on Scheer PAS.

Rapid digitalization of enterprises has improved businesses in numerous ways:

• Increased organizational efficiency
• Reduced the number of individual steps in processes
• Improved the transparency of organizational units
• Provided valuable insights into business processes

… and much more.

However, since the digitalization of any enterprise is a long process that takes time and multiple iterations, the journey into the digital age has often been fragmented. Unable to keep up with the latest technologies and innovations, companies attempt to incorporate different systems, tools, and services to stay competitive.

Because of this, the digital landscape of many modern enterprises now consists of multiple silos—each containing certain systems and services but lacking interconnection or orchestration. Ironically, the very mix of old and new systems meant to bring efficiency, and agility has instead created new hurdles, which seem to grow exponentially over time, making it harder to move forward.

Connecting the Dots

In this new digital age of Agentic AI and the growing risk of bloated digital debt, companies seeking to stay competitive face a choice.

They can “restart” their entire digital landscape with the latest technologies in mind—an enormous undertaking that may take so long that the “newest” technologies are outdated by the time implementation is complete.

Or they can take a different approach: leveraging their existing systems by integrating and orchestrating them into a single modern platform. This strategy not only minimizes downtime and builds on the existing IT infrastructure but also keeps further development simple and streamlined.

How Does Scheer PAS Work?

With the second approach in mind, Scheer PAS, as a digital platform for modern business, enables companies to connect and extend their existing IT landscape with modern technologies while always keeping business processes in focus.

As part of the Scheer Group, which has decades of expertise in business processes, Scheer PAS applies a process-first mindset. This ensures companies never lose sight of what matters most in digitalization: achieving visible outcomes in their core business processes.

With its broad set of capabilities, Scheer PAS empowers enterprises to:

• Create digital versions of their business processes gaining insights into the entire business operation
• Integrate their existing IT landscape (systems, tools, and services) into the platform using premade connectors, adapters, or custom integrations
• Automate processes to eliminate manual work and inefficiencies
• Expose and manage APIs to integrate their services and systems with partners
• Gain constant, comprehensive insights into process efficiency through analytics and data, enabling continuous improvement
• Incorporate Agentic AI into their processes to enhance efficiency, decision-making, and transparency

With these powerful capabilities and its Agentic AI, Scheer PAS, true to the Scheer spirit of business process excellence, enables companies of all sizes to modernize and adapt to evolving market needs.

Backed by success stories from industries such as retail, manufacturing, and public services, along with continuous platform improvements and professional onboarding services, Scheer PAS helps enterprises not only keep pace with competitors but also remain agile and efficient in an ever-changing market.

The post Was ist Scheer PAS? appeared first on Scheer PAS.

The Status Quo of Automation

Especially in the tech world, we can see that many corporations are beginning to replace initial tasks with AI. Looking at today’s market challenges, process automation still largely relies on predefined business rules and fixed workflows to minimize manual work. That model worked in the last century – and up until the late 2000s. But today, faster adaptation is key – driven by two main factors: trends and exceptions.

Anyone who wants to stay resilient must be adaptable. Rigid workflows can no longer cope with the growing number of exceptions, leading to more manual work again. At the same time, trends are becoming more volatile and harder to predict. Large organizations often react with great effort – or simply ignore them. Whether you turn a blind eye to change, like Kodak did, or overinvest in it – the wave will come. Its name is AI. But introducing an LLM alone won’t cut it. What’s needed is a clear concept to create real value – not just the appearance of being “cutting-edge.”

Agentic AI and Process Automation

Just as IT has evolved from machine commands to natural language, process automation must also move forward. The limits of predefined workflows and rules become apparent when dealing with unstructured data and sources. Traditional approaches increasingly require manual input – processes stall, efficiency drops. This is where Agentic AI comes into play. AI agents – powered by secure, locally managed LLMs trained on curated, company-specific data – close the gap between the ambition and the reality of automation.

Where fixed process chains once needed manual checks to handle exceptions, AI agents enable flexible workflows. Examples of what AI agents can balance or enhance within processes include:

• Accessing approved databases and structured data sources
• Comparing information with predefined legal or compliance rules
• Reading and interpreting incoming messages relevant to the process
• Identifying key data and transforming it on demand – without interruption
• Proactively closing data gaps, including automatic requests and feedback handling
• Escalating process steps for human review once risk criteria are met (detected automatically or semi-automatically)
• Ensuring full transparency and analysis at every step – for greater agility and resilience

Looking at today’s Agentic AI features in automation platforms, they primarily replace manual tasks. But that alone doesn’t prevent untapped potential. Even at Kodak, AI agents could have automated many process steps – but without looking at their own data and its meaning, it still wouldn’t have been enough. And that brings us to an aspect we’ve left out so far.

Looking Ahead

The features mentioned mainly reduce manual work and costs. But the real value lies elsewhere: in recognizing improvements and uncovering new opportunities. It’s not just about identifying process errors – Process Mining already does that – but about proactively discovering potential.

When Agentic AI gains insights into business logic, processes, and workflows, it can deliver tremendous value based purely on internal data. What’s often missing is a consistent view of future possibilities. Business analysts can achieve similar results, but often more slowly – and at higher cost.

That’s why it’s worth going one step further: letting AI agents not only replace manual steps but also take over real-time analysis and insight generation. Combined with existing Process Mining, your own data becomes a goldmine – it’s there, just often unstructured.

The Security Question

Anyone integrating AI agents into processes must consider security from the start. Data processing is subject to strict regulatory requirements. Using public LLMs with customer or sensitive internal data increases the risk of leaks and attack surfaces. Responsible implementations account for that from day one.

At the same time, AI agents can be limited to specific data types and volumes – with or without external research. That improves both security and training quality. Clear instructions on how data should be processed and stored increase precision and filter out noise. This way, humans stay in control of decisions rather than being distracted by irrelevant signals.

Conclusion (Part 1)

Back to the beginning: in 1975, Kodak already had all the ingredients for future success. But focusing on existing practices turned a trump card into a missed opportunity. A high degree of automation kept the brand alive for a while – yet the lack of awareness of its own potential contributed to the fall.

Today, anyone automating processes should think beyond cost reduction. Inside every organization lies a data goldmine waiting to be unlocked. Agentic AI helps you tap into that goldmine – boosting both agility and resilience. Stay tuned: In Part 2, we’ll showcase real-world use cases, the Scheer PAS approach to Agentic AI, additional success and failure stories from enterprise settings, and potential risks of over-relying on AI agents. Interested? Subscribe to our newsletter or reach out to discuss your experiences and challenges – together, we’ll find the right solution.

The post Agentic AI: Vom starren Regelwerk zur anpassungsfähigen Prozessautomatisierung appeared first on Scheer PAS.