Observability as code for AI apps with New Relic and Pulumi
Blog post from New Relic
Integrating New Relic's observability platform with Pulumi's infrastructure-as-code enables a comprehensive "observability as code" approach for AI applications, allowing teams to programmatically define monitoring instrumentation, securely manage credentials, and automatically deploy New Relic tools alongside AI infrastructures. This method provides consistent monitoring across environments, facilitates version-controlled observability configurations, and enhances detection of performance issues and insights into AI model behavior. Pulumi offers tools like infrastructure as code (IaC), cloud services, and Environments, Secrets, and Configuration (ESC) for secure management of sensitive data, while Pulumi Copilot aids in generating Python code for infrastructure setup. The guide demonstrates how to implement AI and LLM monitoring in an AI chat application using these tools, with deployment on AWS and monitoring via New Relic dashboards, capturing detailed telemetry data to ensure compliance, quality, and cost control.