cancel
Showing results for 
Search instead for 
Did you mean: 

Conditional Mandatory checks on DFD

KalashSharma
Level 4

Hi All,

I have a requirement is to get PDFs from emails, pass to decipher and then post Decipher verification (if prompted) process them via a BP process. Once the attachments related to an email has been processed by BP, I want to  move that email into a 'completed' outlook folder. Else email should go in 'exception' folder. So maintaining link with the original email is vital in case there are multiple attachments in one email but it is working fine.

So the problem I have is that I can't let user mark an unexpected/wrong PDF as exception or attachment in Decipher verification step. If I let them mark exception in verification step, I (BP process) wouldn't know whether they marked it as exception or it is still waiting for verification completion (as i think the 'get next completed batch' BP action will not report anything related to the exception'ed batch.

So I want to make them select a new field named 'command' on DFD (required, Options= process, Ignore; default=process). When users select the 'Ignore', I would like all required fields on DFD to become 'not-required' so that the PDF can be submitted in Decipher.

I have tried to force the required check using the format regex instead of the 'required' field option so that I can create a a 'specific version' based on command field set to 'ignore' so that I can remove the format regex to make them non mandatory fields. But specific version doesn't get created if there is only a format change. Also I am not sure specific version will solve this even if it did get created as their purpose is different.

what would you recommend to simulate the conditional mandatory check based on a field selected on DFD?



------------------------------
Kalash Sharma
Europe/London
------------------------------
1 BEST ANSWER

Best Answers

Ben.Lyons1
Staff
Staff
Hi Kalash,

There's a lot going on there, so I'll do my best to answer each point.

- Maintaining the link between email and documents. You could add the email details as metadata to the batch. This would mean you have to upload each document as it's own batch. Which isn't really a problem for Decipher, but may make the BP process a little more complex.

- Tracking exceptions. If a document is missing data and can't be processed, you shouldn't override this in Decipher as it will incorrectly update the training dat. It will learn that sometimes those fields aren't populated, which isn't want you're aiming to achieve. You may need to use SQL queries to check which documents (if any) are marked as exceptions. This is dependant internal SQL skills. 

There's an alternative, but it requires some specific updates to your access management. I would recommend requesting an Expert Connect session so we can help guide you through this configuration.

- Specific versions are designed to separate clearly different document layouts (e.g. different vendor invoices) and is managed by the Capture Client prior to data verification. So you won't be able to assign a specific version to a document layout you've already trained (even if data is incorrectly formatted or missing). It also won't be something you can effect during data verification, as the DFD matching/processing has already taken place.

Let me know if I've missed anything.

Also, it really helps us support your query if you can share any screen grabs of your DFD and document samples, where possible.

Thanks

------------------------------
Ben Lyons
Product Consultant - Decipher Specialist
Blue Prism
UK
------------------------------
Ben Lyons Senior Product Specialist - Decipher SS&C Blue Prism UK based

View answer in original post

4 REPLIES 4

Ben.Lyons1
Staff
Staff
Hi Kalash,

There's a lot going on there, so I'll do my best to answer each point.

- Maintaining the link between email and documents. You could add the email details as metadata to the batch. This would mean you have to upload each document as it's own batch. Which isn't really a problem for Decipher, but may make the BP process a little more complex.

- Tracking exceptions. If a document is missing data and can't be processed, you shouldn't override this in Decipher as it will incorrectly update the training dat. It will learn that sometimes those fields aren't populated, which isn't want you're aiming to achieve. You may need to use SQL queries to check which documents (if any) are marked as exceptions. This is dependant internal SQL skills. 

There's an alternative, but it requires some specific updates to your access management. I would recommend requesting an Expert Connect session so we can help guide you through this configuration.

- Specific versions are designed to separate clearly different document layouts (e.g. different vendor invoices) and is managed by the Capture Client prior to data verification. So you won't be able to assign a specific version to a document layout you've already trained (even if data is incorrectly formatted or missing). It also won't be something you can effect during data verification, as the DFD matching/processing has already taken place.

Let me know if I've missed anything.

Also, it really helps us support your query if you can share any screen grabs of your DFD and document samples, where possible.

Thanks

------------------------------
Ben Lyons
Product Consultant - Decipher Specialist
Blue Prism
UK
------------------------------
Ben Lyons Senior Product Specialist - Decipher SS&C Blue Prism UK based

Hi Ben,

Thanks for your quick response. It is helpful.
I was not aware of the fact that submitting some required DFD fields as blanks sometimes would interfere with its learning. And thanks for pointing me in the direction of reading the batch information from Decipher DB layer. It is helpful and I can now design the entire automation in much simpler way.

Regards,
Kalash

------------------------------
Kalash Sharma
Business Analyst
IBM
Europe/London
------------------------------

Perhaps, you could add an action in Decipher BluePrism VBO to get the Excetion'ed batch information within a date range. It will help in many use cases.

------------------------------
Kalash Sharma
Europe/London
------------------------------

Hi Kalash,

No worries.

That's a great idea, please can you submit it as an idea via the "Innovate" button at the top of the page? This will go to our product team for adding to a future release.

Thanks

------------------------------
Ben Lyons
Product Consultant - Decipher Specialist
Blue Prism
UK
------------------------------
Ben Lyons Senior Product Specialist - Decipher SS&C Blue Prism UK based