AI+ Reporting
Updated
AI+ Reporting provides detailed insights into the performance and usage of AI-powered requests across your organization. It enables businesses to monitor, measure, and optimize AI interactions by offering detailed insights at the partner-specific level.
Note: Retention Policy
Reporting data is retained for the last 3 months for each customer and use case. Records older than 3 months will be deleted from the Sprinklr database.
Configure AI+ Reporting
Follow these steps to configure AI+ Reporting:
Step 1: Create a Reporting Widget
Navigate to Care Reporting.
Click the + Create Widget button under a reporting dashboard to build a new report.
For detailed steps, refer to Create a Reporting Widget.
Step 2: Select the Data Source
In the widget setup screen, choose AI+ Reporting in the Data Source field from the dropdown.
Step 3: Configure Metrics and Dimensions
Add relevant Metrics and Dimensions from the Unified Log Reporting source based on what you want to monitor.
Apply filters (optional) to narrow down your reporting.
Supported Dimensions
These dimensions allow you to slice data for deeper insights into AI-based interactions.
Dimension Name | Description |
Workspace | The partner’s workspace where the request was triggered. |
Request Triggered Time | Timestamp when the request was initiated. |
User | The user who triggered the request. |
Audit Request ID | A unique identifier used to track and audit the AI+ request. |
Configuration Set | The configuration set that triggered the request. |
Error Code | A standard HTTP status code that indicates the type of error (for example, 404 for “Not Found”). |
Error | A short description of the issue that occurred during the request. |
Frequency Penalty | A parameter that reduces the chance of the model repeating the same text. |
Presence Penalty | A parameter that encourages the model to introduce new topics in its response. |
Prompt Node ID | The unique ID of the prompt node used to generate the AI+ request |
Provider | The AI service provider used for the request (for example, Azure OpenAI, OpenAI, Google Vertex). |
Provider (Display Name) | The custom display name for the AI provider as configured in AI+ Stud |
Request Completion Time | The exact date and time when the AI+ request finished processing. |
Temperature | A parameter that controls the randomness of the AI response. Higher values produce more creative outputs. |
Supported Metrics
These metrics help quantify AI requests for monitoring and alerting.
Metric Name | Description |
AI+ Request Count | The total number of AI+ requests made during the reporting period. |
Audio Input Tokens | The number of tokens processed from audio input provided to the model. |
Audio Output Tokens | The number of tokens generated when the model outputs audio. |
Cached Input Tokens | Tokens reused from previous requests to improve efficiency |
Image Input Tokens | The number of tokens processed from image input provided to the model. |
Latency | The time (in milliseconds) taken by the model to generate a response after receiving the request. |
Reasoning Output Tokens | Tokens used internally by the model for reasoning before producing the final output. |
Total Input Tokens | The total number of tokens sent to the model as input for the request. |
Total Output Tokens | The total number of tokens generated by the model as output for the request. |
Total Tokens (Input + Output) | The combined total of input and output tokens for a single request. |
Best Practices
Use Partner-Level Reporting for granular insights per partner.
Monitor Latency and Error Codes to optimize AI+ performance.
Track Token Usage to manage resource consumption effectively.
AI+ Reporting enables organizations to gain actionable insights into their AI-driven workflows. By using Partner-Level Reporting for granular analysis, businesses can optimize AI usage, improve operational efficiency, and make strategic decisions.