Amazon Bedrock Monitoring

Monitor Your Models' API Performance and Errors

Centralize your runtime metrics and logs all in one place after installing the Bedrock integration
Visualize the health and performance of your models, including the number of invocations and errors, with Datadog's preconfigured dashboard
View real-time statuses of any monitors configured using Bedrock data, such as high number of invocation throttles or inovcation latency
Easily compare models' latency and performance against one another to find the top-performing solution that best fits your needs

Monitor how frequently customers are using your AI products most and the complexity of data they typically process and generate
Gain insight into optimizing provisioned throughput to balance cost and performance
Verify that your models are being used as intended with input and output token count metrics

Track Bedrock API calls via CloudWatch logs, which is also displayed in Datadog's preconfigured service dashboard
Perform deeper analysis into your customers’ queries to identify common trends and how your models respond
Analyze log data to determine whether the real-world usage of your model fulfills its intended purpose
Understand how to tune your inference parameters for optimizing model performance

Easily create intuitive dashboards— no custom query language required
Track the health and performance of your AWS infrastructure with out-of-the-box dashboards, including AWS service-specific views for EC2, RDS, ALB, Lambda, and more
Receive alerts on essential metrics from AWS services like EC2, Lambda, RDS, and SQS with preconfigured monitors