Hi peers,
I'm working as the Associate Director /Cloud Architect at a Computer Software Company with 10,000+ employees.
Can you please recommend which Process Mining tool (Celonis or UiPath Process Mining) is better for Oracle EBS ERP?
What is the required team size minimum and what should their skillsets be to execute process mining, interpret data, create analysis and simulations?
Overall, how much time does it take to process mining for a typical P2P or O2C process? Is this a one-time activity or continuous improvement program?
Please advise,
Thanks
Hi Nilesh,
Usually, when I need to make process mining I always think in (the most relevant "I guess..."):
1. Which is the application data source (or applications)?
2. How to extract the data (integration services, applications, etc)?
3. How much data do we need to extract?
4. How the raw data is transformed (business rules).
5. How often do we need to run the data mining process?
6. In where would be storage the output data (reporting usability bots)?
7. Is the data confidential?
8. Are there some processes that are in the hold or affect during the process of mining?
9. How many users will be affected?
10. What is the effective time to be ready (after running the process), and consider the SLA Customer?
11. What will be the contingency definition process if the output is not obtained probably?
As mentioned in the lines above the process of mining not only depends of one aspect to consider. In fact when you have integration services ready or direct access to the database "you have the food serve!" and you decide the friendly and trustworthy tool to run the mining process (not only RPA could be handled).
Related to UiPath, I consider it a powerful tool to find, select, extract, transform, export, evaluate, apply business rules and deliver data sets interacting within an application (client-server, web), sometimes when you need to scrap the data from an application you should need more effort because you don't access the information control directly. It perhaps could be a little imprecise and you will spend more time doing and resolving the mining process issues.
Finally, I highlight that the mining process could be improved using scalar architecture design using a bot farm controlled by a load balancer (HTTPS) or network load balancer inclusive using orchestrator only.
"What is the required team size minimum and what should their skillsets be to execute process mining, interpret data, create analysis and simulations? "
I think with an RPA cell that includes 3 persons this task could be managed well, but if you have constraints you need to evaluate the impact. And this definition depends on what quantity of automation needs to be built per sprint.
"Overall, how much time does it take to process mining for a typical P2P or O2C process? Is this a one-time activity or continuous improvement program? "
it's a subjective answer because the size of the organization, the number of reports you need to populate could make the difference. But in general terms, you need to consider that all kinds of mining processes must be certified by functional area, in this scenario you need to consider a continuous improvement program and an RPA cell to support it. More ahead when the bots are in production you need to consider an RPA support cell.
I hope this could be helpful.