Query Document/Dataset (Batch)

introduction while using the docid\ dm3eh gzaocyoqd5y0r8d flow action, it enables to execute a single query on the data of a processed document, using cloudfiles query document/dataset (batch) flow action enables to execute multiple queries on the data of a processed document in a single action what this action does the cloudfiles query document/dataset (batch) action enables you to execute multiple queries (natural language or english language prompts) simultaneously on the data of a processed document the action processes all the queries in a single step and returns the results of all queries together in the action output imagine you are processing kyc files uploaded to contact object records in action using the docid\ dm3eh gzaocyoqd5y0r8d flow action you can classify the uploaded document as a driving license , passport , or neither once classifed, the next step would be to extract all the relevant information, such as the passport number, name, issue date, expiry date etc from the document and then, update the corresponding fields on the contact record classification query prompt "check the type of file if it's a driving license, return 1; if it's a passport, return 2; if neither, return 0 " based on the result (e g , 2 for passport), the flow decides the next steps decision element using the query result, the flow branches into specific paths for each document type follow up queries in a single cloudfiles query document (batch) action for example, if the classification query returns 2 (passport) , the flow enters the "is passport" path and runs queries like "extract the passport number " "extract the name on the passport " "extract the issue date" "extract the expiry date" instead, of using multiple docid\ dm3eh gzaocyoqd5y0r8d actions, you can use a single query document (batch) action and extract results of all the queries in a single output by avoiding multiple query document/dataset actions within a flow and using a single query document/dataset (batch) action you can build complex workflows that extract data from documents efficiently and avoid flow timeouts input parameters let us discuss each of the input parameters in detail, in the sections below agent (optional) the agent parameter determines which ai agent cloudfiles uses when executing a query on a processed document agents define how the query is interpreted and answered if a list of agents was passed in the https //help cloudfiles io/salesforce/process document using ai flow action, then the query document action can use one of those agents to get the response for one or more queries as specified if you specify an agent in the query document action that is not among the document’s configured agents, the query will fail with an error the following screenshot illustrates how to fill in the corresponding input values when a query is executed, the agent is selected in this order agent explicitly provided in the query document/dataset (batch) action if none is specified, the first agent configured at the document level within " process document using ai " flow action if no document level agents are available, the account level default is used this ensures that queries always run with a valid and supported agent, while still giving you flexibility to override defaults where necessary dataset id input the dataset id of the group of files on which you wish to execute the query the dataset id is a unique identifier provided by cloudfiles that represents the data of the collection of processed documents how it is generated ? the id is genrated when a file is processed using the create dataset flow action after this process, a dataset created event is automatically published how to retrieve it ? use the get event details flow action associated with the corresponding event to obtain the id processed document id input the processed document id of the file on which you wish to execute the queries the processed document id is a unique identifier provided by cloudfiles that represents the data of a processed document how it is generated ? the id is genrated when a file is processed using the docid 31ujx1ligtkwfkjuanbzt flow action after this process, a docid\ xrr pnbpocwwcmyg6qgk9 event is automatically published how to retrieve it ? use the get event details flow action associated with the corresponding event to obtain the id alternatively, if you need to execute queries on a salesforce file later or on demand store the file's processed document id in a custom content version field this allows you to query the document anytime, without relying on the event note the same processed document id can be reused multiple times to reference the document and execute queries as needed prompts the prompts input parameter is a collection variable (apex defined type of apex class = cldfs querydocumentprompt) containing all the queries to execute on the file identified by the processed document id steps to configure prompts for each query, create a separate prompt variable 1 create individual prompt variables for each query, create a separate prompt variable example use case extract four details (passport number, name, issue date, expiry date) from a passport document create 4 prompt variables, each of type cldfs querydocumentprompt the number of prompt variables depends on the queries you want to execute 2 assign queries for each prompt variable assign a unique id (used to map the result to its query) assign the query (a plain text prompt specifying the information to retrieve) you can directly pass a string as the query (e g , "extract the passport holder's name " ) alternatively, you can pass the query dynamically through a variable, formula etc allowing to incorporate complex use cases 3 create a collection after defining and assigning all prompt variables create a collection variable of type cldfs querydocumentprompt add all these individual prompt variables to the collection this collection will serve as the prompts input parameter for the action output parameters the action outputs prompts (an apex defined collection variable of apex class cldfs querydocumentprompt) this collection contains all the prompts that were input, along with their corresponding results steps to filter results from outputs 1 loop through outputs use a loop element in your flow to iterate over the prompts collection output by the action considering the same previous example the loop shall look something like this 2 add a decision to check the "id" of current prompt insert a decision element inside the loop to filter prompts based on their id (assigned during input configuration) this helps identify which prompt result corresponds to which prompt query 3 assign the corresponding result as needed after filtering the desired prompt using the decision , access the result field of that prompt variable assign the result to the appropriate field or use it as needed in the flow with reference to the previous example, assign the passport number result to the record field these steps ensure you can efficiently extract and utilize query results from the output prompts collection