AI Action Flows

AI Action Flows act as the orchestration layer between SmartAgent and Amazon Bedrock, allowing AI-powered features such as After-Contact Work automation to be structured, controlled, and customised.

Rather than sending raw contact data directly to Bedrock, an AI Action Flow defines how the information should be prepared, processed, and used. The flow can collect relevant data from the interaction (for example the transcript, customer request, agent responses, or cases created), structure it into a prompt, and then send that request to Bedrock. Once Bedrock generates the output, the Action Flow determines what happens next — such as populating ACW attributes, updating contact records, or triggering additional automation.

In practice, this means AI Action Flows act as the logic and governance layer around generative AI.

Benefits of using AI Action Flows with Bedrock

Controlled AI processing:

AI Action Flows define exactly what data is sent to Bedrock and how prompts are structured. This ensures consistent outputs and prevents unnecessary or sensitive information being passed to the model.

Customisable workflows:

Each organisation can tailor the flow to match their operational processes. For example, different prompts or classifications can be used depending on contact type, queue, or channel.

Integration with existing SmartAgent automation:

The outputs from Bedrock can be immediately used within the platform — such as populating ACW forms, updating attributes.

Consistency and governance:

By standardising how generative AI is invoked, AI Action Flows ensure the same logic and guardrails are applied across all AI-powered features.

Extensibility:

AI Action Flows can incorporate additional data sources or services before or after Bedrock is called. This allows organisations to enrich AI prompts with CRM data, previous interactions, or internal knowledge.

Reduced development effort:

New AI-powered features can be implemented by configuring flows rather than building bespoke integrations with Bedrock each time.

In simple terms, Bedrock provides the intelligence, while AI Action Flows provide the structure and control that turns that intelligence into reliable operational outcomes inside SmartAgent.