Amazon Connect introduces new metrics to enhance AI agent performance

Amazon Connect has expanded its capabilities by introducing eight new metrics aimed at enhancing the performance of AI agents. The newly added metrics include the goal success rate, faithfulness score, and tool selection accuracy, among others. These metrics are designed to provide insights into the quality of AI-driven customer interactions, allowing for the measurement and continuous improvement of outcomes delivered by AI agents.

With these new metrics, businesses can monitor the effectiveness of AI agents in resolving customer inquiries, evaluate the accuracy of responses in terms of faithfulness, and identify any contextual inaccuracies or hallucinations. Additionally, the metrics facilitate the assessment of how accurately tools are selected and utilized by AI agents. Customer feedback can also be captured through thumbs up/down ratings, provided this feature is enabled.

These metrics can be accessed through the AI Agent Performance dashboard within Amazon Connect or via the GetMetricDataV2 API and a zero-ETL data lake, allowing for custom reporting or integration with existing analytics workflows. This new feature is available in all AWS Regions where Amazon Connect AI Agents are supported. For further details, users are encouraged to refer to the Amazon Connect Administrator Guide. More information about Amazon Connect, an AI-native solution that aims to transform customer interactions into memorable experiences, can be found on the Amazon Connect website.