> ## Documentation Index
> Fetch the complete documentation index at: https://docs.open.cx/llms.txt
> Use this file to discover all available pages before exploring further.

# Genesys Cloud summaries, sentiment and evaluations

> Bring Genesys Cloud summaries, sentiment, categories, survey scores and released quality evaluations onto OpenCX sessions for workflows and reports.

Genesys Cloud writes a summary, analyses the conversation and may score it after it ends. OpenCX brings those results onto the session, so you can use them in workflows and reports. Read this to know what arrives, when, and what each needs on your Genesys plan.

## Summaries, analytics and surveys

Genesys takes a few minutes after a conversation ends to write its summary and analyse it. OpenCX checks for them for about an hour after the session closes, then stops, and keeps what it finds on the session for good, so it stays after Genesys deletes its copy. This covers every session linked to a Genesys conversation, calls and sessions from **Show every Genesys conversation** included, but not **Past conversations** imports or conversations that had already ended when they came in.

| From Genesys | On the session |
| - | - |
| Summary | One internal note with the summary (an agent's edit of it, if there is one), the reason, the resolution and the follow-up. The reason and resolution also become `genesys_summary_reason` and `genesys_summary_resolution`. |
| Speech and text analytics | `genesys_sentiment_score`, `genesys_sentiment_trend` (once Genesys has worked it out) and `genesys_categories`, the categories the conversation matched, separated by commas. Topics aren't brought over. |
| A finished survey | The session's CSAT score, unless it already has one. |
| A released quality evaluation | `genesys_evaluation_score`, `genesys_evaluation_critical_score`, `genesys_evaluation_form`, `genesys_evaluation_evaluator`, `genesys_evaluation_agent` and `genesys_evaluation_released_at`, and one internal note with the scores, form, evaluator and agent. The evaluator's comments aren't brought over. See [Quality evaluations](#quality-evaluations). |

Use them like any other session attribute, in workflows and reports. Customers never see the note. With PII redaction on, the summary is redacted before it's saved, and a session whose history was erased gets no summary.

Summaries need `conversation:summary:view`. Sentiment and categories need `speechAndTextAnalytics:data:view` and `recording:recording:view`. Each also needs that feature on your Genesys plan. Surveys need nothing more. If Genesys refuses, **Health** on the **Setup** tab shows **Genesys insights** with what to add, and OpenCX asks again a day later.

## Quality evaluations

Quality evaluations are usually released days after a conversation, so OpenCX looks for them one, three and seven days after the session closes, then stops. Only a released evaluation counts: finished, released to the agent, and not set never to release. If an evaluation is rescored or released again before the seventh day, the attributes follow it, and a new note is added only when the score changed. With several released evaluations, the one released last counts. A session whose history was erased gets none.

Evaluations need `quality:evaluation:view` and quality management on your Genesys plan. If Genesys refuses, OpenCX asks again a day later. **Health** shows **Genesys insights** only if Genesys shows evaluations on your conversations, so an org that doesn't use quality management isn't warned.

## Related

> [card] [CSAT](https://docs.open.cx/resources/csat.md)
>
> How OpenCX skips its own survey while a Genesys one is out.

> [card] [Permissions](https://docs.open.cx/integrations/genesys/permissions.md)
>
> What to add to the role for summaries and analytics.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.