Complaint Triaging Project – ESDC
The Labour Program receives a large volume of complaints from employees. Each complaint is evaluated by an Early Resolution Officer and registered in the system.
Employment and Social Development Canada explains the Complaint Triaging Project from their Labour Program.
Transcript
Slide 2:
The Labour Program:
Works to promote safe, healthy, cooperative and productive workplaces in federally regulated workplaces.
Enforces labour standards legislation and regulations such as minimum wage and hours of work stipulated in the Canada Labour Code.
Slide 3:
The Labour Program receives a large volume of complaints from employees.
Each complaint is evaluated by an Early Resolution Officer and registered in the system.
A Regional Operations Manager assigns the case to a Labour Affairs Officer based on the type.
The Officer investigates the cases until resolution.
Slide 4:
Project goals:
Streamline the complaint assigning process.
Improve the efficiency of complaint assignments.
Expedite triaging the cases.
Slide 5:
We have a solution: A model that supports managers in triaging incoming complaints.
The model detects patterns by sifting through large volumes of internal data from past cases.
Uses a mixture of hard-coded rules and decision trees and makes a prediction on the complexity of incoming cases with an accuracy rate of more than 80%.
Slide 6:
The model sorts the complaints into four bins:
Least complex cases, least likely to result in violations or additional action.
Less complex cases, violations and additional actions unlikely.
More complex cases, violations and additional actions are possible.
Very complex or unusual cases.
Slide 7:
At the regional offices:
Managers can use the bins to:
1- Assign new cases.
2- Identify severe cases.
3- Save time in reviewing incoming complaints.
Slide 8: Next steps:
Continue to work with our internal partners to conduct a controlled experiment and measure the impact of the triaging tool on process efficiency.
Thank you.
Stay tuned.
Resources/ Courses
· Discover Series: Discover Data