One area for improvement is that, while Celonis has existing APIs for some domains like accounts, it lacks ready-made ones for others, like financial services. This means we need to customize our data channels so the data can be used by Celonis. There's a lot of pre-work needed to cleanse and harmonize data attributes before it's usable with Celonis. That's the main issue I see with the tool. Let's use an example of a financial process, such as loan origination. An application comes in and is scanned. Someone then reviews it, capturing the details. After review, it enters the organization system, potentially with bypass steps. There are different stages: validating the application, capturing data, validating documents, and deciding whether to accept the loan. We need to precisely define these stages. In many organizations, this data might not readily exist. We can't just use Celonis with existing data; it needs harmonization first. This means we might not be able to leverage existing data in Celonis, starting from scratch instead. If an organization has ten years of data, we need to cleanse it at the data point level and metadata level. This involves merging, simplifying, removing, and relabeling attributes. It's a lot of work. Celonis works well when this pre-work is done. Right now, good facilitating tools are missing, even from Microsoft. We need to harmonize those timestamps manually before connecting the data to Celonis, which provides the analysis output we need. I hope Celonis simplifies this process in future releases.