How come this setting causes a validation failure when importing deflection articles?
We are currently executing a migration from Zendesk Guide to Genesys Cloud CX, focusing on the AI Bot deflection capabilities. In Zendesk, we relied heavily on our knowledge base articles to deflect simple queries before routing to agents. The goal is to replicate this behavior using the Genesys Cloud Bot deflection settings.
Environment Details:
- Platform: Genesys Cloud EU1
- SDK: Python 3.9 with Genesys Cloud Python SDK v2.5.0
- API Endpoint:
POST /api/v2/bots/{botId}/deflection - Source Data: Exported Zendesk Guide articles (JSON format)
The specific issue arises when attempting to programmatically create deflection entries. While the Zendesk articles are structured as simple markdown with title and body, the Genesys Cloud Bot API seems to require a more complex interaction model definition. When sending the payload, the system returns a 400 Bad Request error. The error message states: "Validation failed: One or more fields contain invalid data. Field: 'content'. Message: Content must be a valid interaction or script reference."
In Zendesk, the content was just text. In Genesys, it appears the bot expects a reference to an existing interaction flow or a specific script ID, rather than raw text content. This is a significant shift from the Zendesk model where the article itself was the answer.
I have tried mapping the Zendesk article ID to a Genesys Cloud Script ID, but the validation still fails. It seems the bot is looking for a structured interaction node rather than a simple text block. Has anyone successfully migrated Zendesk deflection content to Genesys Cloud Bots? What is the correct payload structure for the content field when the source is a static knowledge base article?
How do I correctly format the deflection content payload to map Zendesk Guide articles to Genesys Cloud Bot deflection without triggering a 400 validation error?