I am currently utilizing the DU template for a classic project. I will be using the main-ActionCenter.xaml workflow because my process will be unattended.
I have annotated in Data manager and created a new extractor in DU cloud. I am unsure what the next step will be to continue in the studio flow, can you provide a step by step for this please. I started looking at the studio flowchart but it requires “getTransactionItem” from a queue. I have created that queue, but all of my files are PDF and the queue only takes csv.
→ You have to send the pdf file paths as a value in the queue item. Store the path of the pdf files in a list variable and iterate through each item and add the path of the file to the queue.
→ You need to provide the ML Skill and its endpoint, which you obtained from the AI Center after training your ML model, to the config file. Then, open the template, navigate to the ML Extractor, and call the ML Skill and its endpoint from the config file.
→ Mention the mandatory and non mandatory fields in the invoice processing sheet if you are working with invoices.
→ Change the classification threshold values as per your requirement.
→ Configure the classifiers and extractors in the classification scope and extraction scope.
→ Make changes in the extraction validation to check the confidence score validation and other required validation on your requirement.
It’s better to go through each state and each workflow in DU framework and don’t forget to look into Config file which plays a key role here.
Hi @mkankatala , thanks for this. For the first step, would I have to make a seperate process to add the items into queue, because I cannot directly update them into the queue in cloud
Yes you can build a different process to add the data to the queues, we will generally call it as dispatcher… @ashton.lalchan
If you are aware of AI Center you will create a Project in the AI Center, In that project you will create datasets, importing the Invoices or Files to it and do mapping for those files (Training) in the Document Manager, Creating the ML Package, Running the Pipelines, Creating the ML Skill after creating the ML Skill you will get the two parameters one is ML Skill and other one is Endpoint.
You will use both values in the ML Extractor activity.
If you want to learn about Document understanding and AI Center we recommend you to visit the UiPath Academy and search for the courses, you can grab the whole knowledge from there.
Anil_G
jast1631
yuichi