Genesys Cloud - Main

Hello everyone,

For projects with many Data Actions, what techniques have you used to improve performance?

I'm particularly interested in:

• Reducing API latency
• Avoiding unnecessary requests
• Caching strategies
• Best practices for integrations

Any recommendations would be appreciated.

Hi Fausto,

Great question.

From my experience, the biggest performance improvements come from reducing how often external services are called rather than trying to optimise Architect itself.

A few practices that have helped:

• Call a Data Action only when the result is actually needed. Use decisions early in the flow to avoid unnecessary requests.
• Reuse data already returned from an earlier lookup instead of calling the same API again later in the interaction.
• Where possible, combine related lookups into a single backend request rather than using several sequential Data Actions.
• Keep request and response payloads small and return only the fields Architect needs.
• Set sensible timeout and failure handling, with a fallback path so the customer journey does not stall when an external service is slow or unavailable.
• For relatively static data, use Architect Data Tables or caching in the integration layer rather than calling the source system on every interaction.
• Test each Data Action independently and monitor response times so it is clear whether latency is coming from Genesys, the network, or the external system.

In our integrations, avoiding duplicate lookups and keeping the external API response focused on only the required fields has made the biggest difference.

In my very limited experience adding delays or loops for populating data (mostly with real time Genesys APIs) is unavoidable.