#Einstein For Service8 diskutieren mit

@Tugba Salis

your read is correct here - this is a hard limitation, not something that you are missing in the config.  

Standard capability templates have a fixed input defined by the feature that invokes them. Einstein Email Summaries fires against the email itself, so the capability hands the template the email content and nothing else. 

So the way I see it you have two options here:  

Option 1 - you may be able to parse the Case Id from the email text. If your org uses ref ID threading for Email-to-Case, then the case id would be in the email body, and you can use the prompt flow to parse it out.  

Option 2 - build a Flex template instead. Flex templates let you define your own inputs, so you can pass the Case record and use it how you please. You can use this through a quick action / screen flow / agent action / apex - whatever works for you. 

Let me know if that makes sense  

Thanks 

Steve 

Hi @Prashant Jain

 This looks like an onboarding/provisioning issue rather than a tenant-key issue. The error 

“Organization is not yet onboarded to GDoT”

generally means the Salesforce org needs to be provisioned/onboarded before the Anypoint Platform connection can be established.  

Please check with your Salesforce/MuleSoft admin or MuleSoft Support/CSM team to confirm the org's GDoT onboarding status. Once onboarding is completed, retry using the Salesforce Org Tenant Key from Setup → MuleSoft → Anypoint Platform Setup  

https://docs.mulesoft.com/access-management/connecting-salesforce-orgs?utm

 If you find this response helpful, please mark it as the Accepted Answer, as it may also help other Trailblazers.  

Two things to line up for dynamic choices from a Flow in an Enhanced Bot: 

Records not showing - almost always the variable types or the output mapping: 

- In Bot Builder, the variable receiving the list must be type Object with 'Contains List' checked (a record collection). If it is Text, an SObject list will not map and you get nothing. 

- The Flow's output variable must be a record collection (List of Lead) AND marked Available for Output. Then in the bot's Flow action step, map that Flow output to your Object-list bot variable. A missing or mismatched mapping here is the usual output-invocation-mapping issue and shows as empty choices. 

- Sanity-check the Get Records actually returns leads (filters / running-user access). 

Customizing the choice label - that is set in the Question's Dynamic Choices config, not the Flow: after you point it at the Object-list variable, you choose which field is the Label (what the button shows, e.g. Name) and which is the Value (e.g. Id). For combined text like 'Name - Status', add a Formula or Text field in the Flow, include it in the collection, and pick that as the Label. 

So: Object plus List variable on both sides, output marked Available for Output and mapped, then set Label and Value in the choice config. 

If this helps, please mark it as the Best Answer so it helps the next person - thanks 🙂

Hi @Bhavin Patel, what are businesses using for reporting on Einstein Case Classification? We are looking to understand the accuracy of the application in predicting field values (including if there is a percentage we can show) and how many times service agents are changing the value predicted by the application. 

Any advice would be much appreciated! 

Thanks,

Ali