Timeline Process Intelligence – an Academic Evaluation

I used TimelinePI in my research for my Bachelor Thesis with the intent to test and evaluate six different process mining software utilities with a set of event logs.

In my evaluation, one of the unique functions of TimelinePI which I have come to appreciate is, the unique representation of process paths as timelines instead of a process model, also known as process schemas. Unlike other vendors, which display the process cases as a process model in nodes and paths which can only sort the cases by duration and number of events, TimelinePI was able to identify cases which have the longest throughput times and gives more granular detail into the actual process execution as it occurred. Additionally, the ability to filter and group timeline variants was very practical to show the most frequent paths in the process model and zoom in on process details, allowing me to scope a detailed analysis.

One advantage of TimelinePI was the ability to make a direct comparison, including an animation, of two different process variants, so that you could directly point out the differences of two different executed paths. This was practical for performing a variance analysis of a process path against the most frequent process path or against the “happy path”.

Another great function I found with TimelinePI was the ability to query the process model selectively for loops. With this type of search ability, it was possible to identify specific cases and easily expose activities which contain direct or indirect repetitions. The ability to show paths of processes and the activities performed in the form of arrows (Path Analysis) proved to be very practical to finding steps that were repeated, skipped, or done out of order.

Unlike other process mining software vendors, TimelinePI offers the function to automatically choose milestones in the recognized process model algorithmically. By using advanced statistical analysis, the software chooses which steps are most reliably and commonly occurring at the same place in the process automatically.

The ability to define your own KPI criteria such as the maximum duration of an activity, the ideal path and the cost per activity is unique to TimelinePI. Additionally, TimelinePI offers the possibility to trigger alerts/warnings which can occur on specific conditions and continuously monitors processes. For example, if an activity has become overdue the system can send a text message to let someone know or trigger an IT system to remedy the issue in real time. This allows for constant real time process oversight and compliance monitoring.

Overall the TimelinePI platform was easy to learn and use. The software was intuitive and helpful for my analysis and my evaluation. I appreciated how simple it was to craft different analyses and visualizations. Almost every type of analysis I needed to perform was preconfigured and created automatically by the platform. As the Timeline platform runs in a cloud infrastructure there was not any limitation in uploading size of the event logs. This made it possible for me to analyze data sets with over 1 million entries for activities. Another great option of TimelinePI is that there is the possibility to run an unlimited number of projects at the same time, as well as there is the possibility to upload process data from numerous sources including the ITSM tool ServiceNow through an API. Additionally, for someone not familiar with common ETL function, TimelinePI offers ETL tools in the cloud, so you can load raw data and perform transformations in the cloud and loud the data directly into your projects.

If there were one thing I thought that TimelinePI could improve on is a feature relevant to Social Mining, or resource mining. An example of this would be creating a process model from the resource perspective (Social Net), the most frequent resource in the whole process model or doing a bottleneck analysis from resource perspective (i.e. Which resource had the most working time/waiting time?).

Note from the TimelinePI team: this type of feature is in development. We continue to listen to customers and the process analytics community to develop and grow the Timeline Process Intelligence Platform. Feature requests can be sent to support@timelinepi.com or submitted at support.timelinepi.com.

In general, I recommend TimelinePI for business users who want to set their focus on the identification of specific cases, the visualization of process execution, or want to quickly and easily find bottlenecks, process problems, and improvement opportunities within their processes. Timeline Process Intelligence is also great for users who don’t want to code or do not have the coding skills, as the platform offers standardized point and click analysis tools to allow users to easily scope and define their own analyses and KPIs without the need to code or have advanced technology experience.

About the author

Christian Erling graduated with a degree in Business Computing from the University of Passau, Bavaria, Germany. His research has been focused deeply in Process Mining and the capabilities of Process Analytics software. Christian has been working continue researching in process and business computing within the Department of Data Science and Distributed Information Systems at the University of Passau as Java Developer. He will continue to focus later in his Master studies in business computing especially in Process Science as well as in Software Development. Connect with Christian on Linkedin.

Christian acquired a TimelinePI Academic License and used TimelinePI for his research at the University of Passau, Bavaria, Germany. If you are interested in acquiring an Academic License, or learning more about becoming an Academic Partner visit the Academic Partner Alliance page and get started today.

February 18, 2019