Series: Crear un agente copiloto de Genesys

Best practices for your Agent Copilot initial deployment

To ensure a smooth and effective launch, please complete the following steps ahead of the official kick-off:

  • Take natural language understanding (NLU) first steps: Begin with simple trigger responses to support agents at key conversation points, such as displaying a welcome greeting at the start and providing predefined (canned) responses at the end of the interaction.
  • Refine knowledge base: Review and update key knowledge areas to ensure accuracy and relevance.
  • Enhance dictionary management: Add relevant keywords and phrases to the dictionary to improve transcription accuracy. During the first 2–3 weeks, focus on refining the dictionary based on agent and team leader feedback. For more information, see Work with dictionary management.
  • Involve team leaders: Team leaders should closely monitor agent interactions, provide regular feedback, and guide agents in transitioning away from previous note-taking habits.
  • Allow time for transcription improvements: Expect transcription quality to improve gradually over the first few weeks as the directory management is being built out.
  • Track feedback and adoption: Monitor feedback and capture examples of how agents benefit from AI-generated notes.
  • Support agent adjustment period: Allow agents time to become familiar and confident with the AI note-taking system.
  • Configure agent copilot by queue: Tailor the Agent Copilot setup to the needs of each specific queue for optimal performance.