Process Mining meets BPM

The acquisition of Symbio marked an important milestone in the journey of Celonis - it’s adding critical Process Management capabilities and is thus a central piece of comprehensive Process Intelligence. It enables a much broader thinking about processes and at the same time reinvents how the foundation - process modeling - works.

But one step after another: Essentially, we’re talking about bringing Process Mining and Process Management (BPM) together. While both have their own use cases and capabilities such as Process Discovery, Analytics, Monitoring, Exploration, Alerting in the Process Mining space and Business Modelling, Process Planning, Documentation (i.e. for QM or ITSM certifications), Employee rollout and onboarding, etc in the Process Management space, there is a central thing where the two worlds come together: The Process Models.

And here we see two sides of the coin: The as-is (as discovered through Mining) and the to-be (as planned or documented in BPM). Although we’re talking about the same coin, there is typically a discrepancy. Let me share a few insights about how we bring those two sides together, what the role of AI is, and how to generate operational value for an organization with that.

Comprehensive Process Intelligence Graph

Process Mining provides clear insights into how the business operates, how processes are executed operationally, what bottlenecks and value opportunities do exist. It generates a digital twin that well reflects the operational reality, that uncovers improvement potentials and is capable of quantifying it. Process Management now enriches this digital twin by providing further context: It embeds the observed processes into a broader process landscape, connects them with to-be states and adds important information that are critical to take into consideration when thinking about processes. Finally it connects processes to all business relevant dots such as people, technology and risks which is a prerequisite for getting into the driver seat for any transformational change.

Process Modeling reinvented

Traditionally, capturing business processes is a hard manual effort to get in-depth knowledge enabling you to generate proper documentation, rethink processes and continuously improve them. Instead of doing this from scratch, there are two new ways to get there:

GenAI-based modeling

Maybe you have tried popular AI technologies like ChatGPT on your own already. The better the input (the prompt) is, the better results you will get. In BPM, you have a lot of context available by design which can be used for that prompting. Let’s say you want to design a new process for running digital campaigns in the UK considering data protection regulations (such as GDPR), that considers the best working channels and formats and takes best practices from other regions of your company into account. And now imagine the system creating the model for you instead of you talking to x people to get the knowledge in y workshops and then consolidating this in a process model. That’s not just imagination but reality in Celonis’ Process Designer already today.?

Mining-based grounding

If you’re already running a process and need proper documentation, you would traditionally start talking to SMEs, LoB owners and the people that are executing it. You will get to different viewpoints, mixing the “should-be” and the reality and biased, e.g. because important practices are hidden. Think about an alternative which is grounded in the data. Process Mining is used to get there; you pick execution variants that you consider as good, add the context (like responsibilities, risks, etc.) and change it where necessary to turn it into a reference model - all in one single user experience. That’s not only much faster and rooted in reality, but it also delivers a solid foundation for adherence checking, since you now work on a level of detail you have proper data points for.?

Living process models

Many organizations face the issue of “dead” process documentations - means processes have been documented, but are not consumed, not updated and thus, not relevant for people's daily business. Does that mean documentation is worthless? Clearly not! But to generate value, transparency and attraction (in a sense of making them relevant for people's business reality and a joy to work with), those process models must live. Celonis Process Management solves this by augmenting transactional data, gathered from Mining, on Process Models so that quantities, performance metrics (KPIs), errors and automation rates etc. are fully transparent at any time. We call this the Process Cockpit as it does not only provide institutional guidance to the process model consumers but also the view outside so that they can navigate the real world.

Process Adherence

One key challenge of Process Mining is interpreting what you see in a sense of differentiating good from bad behavior, making statements about whether SLAs and quality promises are met and, as a consequence, detecting the right projects that really move the needle. In essence, you need a way to ensure processes are executed under adherence to best practices and (quality and industry) standards in the most efficient way. This is what we call Adherence - bringing mining, modeling, conformance and compliance together instead of working with each piece in isolation. With that, the Process Adherence Manager is the backbone of a structured and steered process that aligns with the organizational objectives, fosters efficiency and reduces risks. It enables users to identify value opportunities comparing their event data against a to-be definition (see above the Mining-based grounding) of how the process should run. Instead of you looking for insights, the system tells you where processes in reality deviate from this to-be and what happened instead. The significance of each deviation is automatically assessed by their performance impact, which makes it easy to spot the opportunities with the highest value. And finally, the value realization is tracked in the Transformation Hub. For the first time, there is an end-to-end approach ranging from process design over adherence to a value realization that can be quantified in data.

This shows well how valuable the capability to navigate in the intersection between Process Mining and BPM is. However, this is only the beginning. We truly believe and already work on making this combination even stronger. Just think about what would be possible when you really used all your process data to create better and more accurate models. Or when you had all contextual information from BPM available for Mining. There is a lot to come, so stay tuned!