Work with advanced custom summaries

AI-generated summary

Genesys Cloud's Agent Copilot streamlines interaction documentation through automated conversation summarization with customizable AI-powered capabilities. The system supports immediate post-interaction summaries through AI Studio's Summaries feature, requiring Genesys Cloud CX licenses and Agent Copilot admin roles. Customization options encompass tone, length, structure formatting, focus areas, perspective choice, and entity handling, allowing organizations to tailor summaries to their specific needs. The platform accommodates various structures, including subject information, participant details, call context, resolutions, and action items. Implementation best practices emphasize clear model context, specific content requirements, neutral language, consistent formatting, and efficient prompt design to minimize latency. The configuration process utilizes AI Studio's Summaries feature, accessed through the Orchestration menu, where users can create and test custom formats. components must capture key topics, descriptions, resolutions, action items, and relevant entities such as dates, policy numbers, and amounts. The system particularly supports financial services use cases, including third-party call tracking and upsell opportunity documentation, while maintaining structured formatting with bullet points, consistent spacing, and first-person language options

PrerequisitesClick to expand

Genesys Agent Copilot’s summarization feature provides immediate conversation summaries after interactions. This functionality enables more accurate interaction capture and faster review by supervisors or quality managers and allows agents to focus on customer experience.

You can use custom prompts to personalize how Genesys Agent Copilot creates the summaries of agent–customer conversations. You can define the tone, length, structure, and focus of your summaries, for example to emphasize customer sentiment, reveal key issues, or produce concise overviews.

Best practicesClick to expand

Define who the summary model is and what their specific task is to ensure that Genesys Agent Copilot understands the context and purpose of the prompt.
Describe how you want the summary structured, such as a paragraph or sections, such as Issue, Actions, Resolution.
Specify whether the summary is written in first, second, or third person.
List exactly what must be included (such as issue, actions, outcome) and what must be excluded (such as greetings, verifications).
Instruct the model to reproduce all entities (such as names, dates, or IDs) exactly and to leave them blank if it finds no relevant data.
Indicate whether you want a short summary or a more detailed one.
Tell the model not to infer or add information that is not explicitly stated in the transcript.
Ask the model to maintain a neutral, factual voice without emotional or subjective language.
Use compact, direct sentences and remove unnecessary words, symbols, or formatting to stay within the character limit.
Specify which aspect of the conversation must receive emphasis, such as actions taken, resolution, or customer impact.
Use uniform wording, structure, and style across all prompts, which helps ensure consistent model behavior and evaluation.
State each instruction only once to reduce redundancy and prevent conflicting guidance.
Keep the prompt simple. Use straightforward language and avoid special characters, extra spaces, parentheses, or dashes that can confuse parsing.
Clearly define naming conventions, such as referring to participants as the client and the agent for consistent summaries.
Because they can increase latency, avoid long prompts.

Sample promptsClick to expand

Summarize customer service transcripts into structured sections: Subject, Participant Info, Call Date/Time, Reasons, Topics, Next Steps, and Entities (with values listed below). Exclude PII and avoid adding facts not present. Keep the language concise, factual, and neutral. Voice transcripts can contain errors. Summarize key details only.
Summarize the conversation step by step into sections: Overview, Issue, Agent Actions, Resolution, and Entities. Use clear line breaks and factual, professional language. Exclude greetings, personal data, and assumptions. Reproduce entities exactly as stated, leaving blanks if not mentioned.
Summarize this conversation focusing on actions taken and any policy or compliance aspects discussed. Use sections: Subject, Policy Context, Actions, Outcome, Entities. Exclude greetings and small talk. Copy all entities as written, leaving blank if none. Keep the tone formal and factual.
You are a call center for a financial, retirement, investment, and insurance company. Summarize agent–customer interactions succinctly without adding facts. Refer to yourself as “I.” Use sections: synopsis, reason, resolution, action items, and entities (with values or blank). Keep PII visible. Write concise bullet points, first-person language, clear spacing, and varied, non-repetitive wording.
You are a car finance institution. Summarize calls using the customer’s first name. Include sections: key topics & descriptions, resolution, action items, and entity list (Dates, Policy Numbers, Policy Types, Amounts, leave blank if missing). Note if someone calls on another’s behalf or if upselling occurred. Exclude ID verification. Keep spacing clear and language varied, not repetitive.

Create a custom summary prompt

Click Menu > Orchestration > AI Studio > Summaries.
Select a summary from the list or create a new summary. The summary editor appears.
Under Advanced configuration, click Advanced configuration. The Advanced configuration window appears.
Enter a prompt in the Custom prompt box.
Click Save.
Test your configuration. For more information, see Test a summary configuration.

[NEXT] Was this article helpful?

Get user feedback about articles.

Work with advanced custom summaries View summary

Create a custom summary prompt

[NEXT] Was this article helpful?

Work with advanced custom summaries