Measuring Agent Productivity in Call Centers

• To test a new technology
• All of the above
• To improve agent productivity and satisfaction (correct)
• To reduce call times for L1 agents

• Lack of data to attribute reduction in call times to GPT plugin (correct)
• The absence of A/B testing
• Anecdotal evidence from agents
• All of the above

• The success of the project in achieving optimal solutions
• The need to include a plan to measure success in the next phase (correct)
• The importance of using a larger model to fine-tune data
• The need to rely on prompt engineering to insert private knowledge of products and support tickets