Hello everyone! I am a Genesys PS consultant and I am currently designing an automated quality evaluation workflow for a large client. We want to implement a system where a certain percentage of evaluations are automatically selected for ‘Calibration’ every week. I am looking for a way to use the Quality Management API to identify evaluations that have a high variance between the agent’s self-score and the supervisor’s score. Does anyone have a script or a Data Action that can pull these scores and flag them for a calibration session?
Hello Emi24. I am a DevOps engineer and I manage our configurations using Terraform. You can definitely automate this using the /api/v2/quality/evaluations/query endpoint. You should write a Python script that pulls all completed evaluations for the week and calculates the absolute difference between the agentScore and the totalScore. If the difference is greater than ten points, your script can then use the API to move that evaluation into a specific ‘Calibration’ queue. It is a very efficient way to ensure your grading is consistent across the entire organization!
Hey Emi24. I am an admin for a large organization in Japan. We tried to do this with a Data Action but we found that the evaluation query API has a very low rate limit for bulk requests. If you have thousands of agents, your script might take several hours to run. We ended up using the ‘Analytics Conversation Detail’ API to find the interactions first, and then we pulled the evaluation data only for the interactions that met our criteria. It was much faster!
Greetings. I am also an admin in Japan. To follow up on Sor55, please make sure you are handling the ‘Calibration Object’ correctly in the API. When you flag an evaluation for calibration, you should create a separate calibration record so that the original evaluation score is not overwritten. This allows you to track the ‘Calibration Delta’ in your reporting dashboards, which is a key metric for our quality team.