Fusion People Hub

Introduction

In many Oracle Fusion HCM implementations, performance documents contain ratings from multiple sources: peer feedback, manager feedback, behavioral KPIs, business KPIs, and overall summary sections. While these ratings may be captured separately during the review process, organizations often need a consolidated score that can be calculated consistently and written back to the performance document in a supported HCM Data Loader format.

This blog explains a practical end-to-end solution for that requirement. The approach uses a BI Publisher report to calculate the final score, prepares the output in HDL-supported format, bursts the generated file to Oracle WebCenter Content, and uses fast formula logic to transform the report output into the required HDL business objects. The solution is aligned with Oracle’s supported Performance Document HDL capability, where the ORA_RATING_PD operation is used to provide section ratings and comments for existing performance documents.

Business Requirement

The business requirement is to calculate the total score for a performance document by combining questionnaire-based peer feedback and manager-driven performance ratings. Once the score is calculated, it must be translated into the correct rating tier and loaded back into the employee’s performance document, specifically into the Overall Summary section.

The expected output should contain the employee person number, assignment number, manager person number, performance document name, final calculated score, section information, and rating name. This output is then converted into an HDL file that updates the RatingsAndComments component of the PerfDocComplete business object.

Solution Overview

The solution is divided into four major parts. First, a BI Publisher data model identifies recently updated performance evaluations and calculates the required scores. Second, a bursting data model sends the generated text output to UCM using the predefined content server delivery channel. Third, an HCM Data Loader fast formula maps the report output into HDL metadata and merge lines. Finally, a flow schedule formula runs the process at a defined interval so that rating updates can be automated.

BI Publisher Report –

Fast Formula for HDL Transformation

Fast Formula for Flow Schedule

Reference HDL Output Format

The generated HDL file contains two sections. The first section updates the performance document using the PerfDocComplete discriminator. The second section updates the rating details using the RatingsAndComments discriminator. The rating name is populated from the tier description derived in the BI Publisher data model.

METADATA|PerfDocComplete|AssignmentNumber|CustomaryName|Operation
MERGE|PerfDocComplete|E506509|OE - Annual Final Review 2025-26|ORA_RATING_PD
METADATA|RatingsAndComments|CustomaryName|AssignmentNumber|ParticipantPersonNumber|ParticipantRoleTypeCode|SectionName|SectionTypeCode|RatingName
MERGE|RatingsAndComments|OE - Annual Final Review 2025-26|E506509|506584|Manager|Overall Summary Section|OS|3 - Successful

Payroll process flow

The payroll process flow acts as the orchestration layer for this automation. Instead of asking users to run the BI Publisher report, download the output, upload the file, transform it, and manually submit HCM Data Loader, the flow pattern connects all these steps into one scheduled process. In this solution, the flow refers to the schedule formula, the HCM Data Loader transformation formula, the BI Publisher report path, and the upload/load tasks required to move the calculated rating data back into Oracle Fusion HCM.

At a high level, the flow follows a generate-transform-load pattern. The BI Publisher report first generates the source data and bursts the output to Oracle WebCenter Content. The payroll flow then uses the generated Content ID as the input for the transformation step. The transformation formula converts the pipe-delimited report output into an HDL-compliant file, and the final task initiates HCM Data Loader so the rating information is merged into the performance document.

Step 1: Create the Custom Payroll Flow Pattern

The first step is to create a custom payroll flow pattern that will control the full automation. This flow pattern should include the tasks required to run the BI Publisher report, generate the HDL-ready file, initiate the data load, and optionally produce an error report. The flow should be created under the appropriate Legislative Data Group so that it is available to the right payroll and HR operations users.

Step 2: Add the Generate Data Task

The generate data task is usually based on the Run BI Publisher Report payroll flow task. This task points to the custom report path created in BI Publisher and executes the report as part of the flow. The report output should be generated in text format and delivered to UCM through bursting. The report path and report name must be correctly maintained in the task parameters so that the flow can locate and run the report without manual intervention.

In this blog’s example, the report generates the performance rating data and creates a file such as PerfDocComplete.dat. The bursting setup writes the file to UCM using the configured delivery channel, security group, and content title. This makes the output available for the next payroll flow task.

Step 3: Use the Schedule Formula

The schedule formula controls when the payroll flow should run next. Instead of using only the standard scheduling options such as once, daily, weekly, or monthly, a flow schedule fast formula can return the next scheduled date based on custom logic. In this implementation, the formula adds approximately five minutes to the previous scheduled run time, allowing the process to check frequently for recently updated performance documents.

This is especially useful for near-real-time automation. Whenever a performance evaluation is updated, the next scheduled execution can pick it up, calculate the final score, and push the rating back through HDL without waiting for a manual batch process.

Step 4: Generate the HCM Data Loader File

After the BI Publisher report output is available in UCM, the next task is to generate the HCM Data Loader file. This step uses the source Content ID and the HCM Data Loader transformation formula. The transformation formula reads the source file line by line and converts each input row into the required HDL metadata and merge rows.

For this solution, the transformation formula produces the PerfDocComplete parent row and the RatingsAndComments child row. The parent row identifies the performance document and operation, while the child row updates the participant, section, section type, and calculated rating name.

Step 5: Initiate HCM Data Loader

Once the transformed HDL file is generated, the flow initiates HCM Data Loader. This task takes the generated HDL Content ID and submits the import and load process. HCM Data Loader validates the file, applies business-object validations, and merges the rating information into the target performance document.

The flow should also include an appropriate process configuration group. This allows administrators to control processing behavior such as logging, thread count, and error mode. These settings are important when the automation runs frequently or processes multiple employee records in a single execution.

Step 6: Add Upload and Content Handling Logic

The file movement between BI Publisher, UCM, and HCM Data Loader depends on accurate content handling. The report bursting setup should create a unique content title or content ID for each run. This avoids overwriting previously generated files and makes it easier to trace a payroll flow execution back to the report output it produced.

The Call ID data model supports this traceability by retrieving the ESS request call ID and using it in the bursting logic. When troubleshooting, this linkage helps identify which report output, UCM file, transformation run, and HDL load belong to the same scheduled execution.

Step 7: Build Error Handling and Monitoring

Error handling is an important part of the payroll process flow because a payroll flow task may complete successfully even when the downstream HDL load contains rejected records. To make the process operationally reliable, add an error-reporting step after the HCM Data Loader task. This report can read the HDL load results and display rejected records, validation messages, and failed business-object rows.

For support teams, this provides a single place to review the outcome of each run. Instead of checking multiple ESS jobs manually, users can open the payroll flow results, review output files and logs, and confirm whether the rating update completed successfully or requires correction.

Recommended Flow Sequence

1. Generate Data – Run the BI Publisher report and create the source text file.
2. Burst Output to UCM – Store the generated report output in WebCenter Content.
3. Generate HCM Data Loader File – Use the transformation formula to convert the source file into HDL format.
4. Initiate HCM Data Loader – Submit the generated HDL file for import and load.
5. Trigger HDL Error Report – Review load failures, rejected rows, and validation messages.
6. Schedule Next Run – Use the flow schedule formula to calculate the next execution time.

End-to-End Process Flow

1. The scheduled flow runs at the configured interval.
2. The BI Publisher report identifies recently updated performance evaluations.
3. The main data model calculates questionnaire and overall summary scores.
4. The final score is mapped to a rating tier.
5. The report template generates a pipe-delimited HDL-ready text file.
6. The bursting query sends the file to UCM using the WCC delivery channel.
7. The HDL transformation formula converts the text output into PerfDocComplete and RatingsAndComments records.
8. HCM Data Loader processes the file and updates the performance document rating.

Conclusion

This approach provides a repeatable and automated way to calculate performance document scores and update final ratings in Oracle Fusion HCM. By combining BI Publisher, bursting to UCM, HDL transformation fast formula, and flow scheduling, the process reduces manual intervention and keeps performance ratings aligned with the organization’s scoring rules.