Coming soon
OpenTelemetry ingestion for SpendLens.
Connect existing OpenTelemetry instrumentation to SpendLens and turn standard AI traces into workload-level cost attribution and engineering actions.
This integration is not available yet.
Python and Node.js SDKs are available today. OpenTelemetry ingestion is planned for teams using Java, Go, .NET, and other instrumented services.
What it will enable
- Send standard OpenTelemetry traces or spans to SpendLens.
- Attribute model, token, latency, cache, and cost data to services and workloads.
- Use existing OpenTelemetry Collectors and observability pipelines.
- Keep AI provider requests on their current direct path.
Who it is for
Platform and engineering teams that already instrument Java, Go, .NET, or mixed-language systems and want SpendLens insights without adopting a native SDK in every service.
Planned telemetry
Planned support includes standard generative-AI provider and model attributes, token usage, latency, errors, service identity, deployment environment, and workload context. Final field support may change before release.