Claus Ibsen created CAMEL-23861:
-----------------------------------
Summary: camel-langchain4j - Add OpenTelemetry span attributes for
AI/LLM observability
Key: CAMEL-23861
URL: https://issues.apache.org/jira/browse/CAMEL-23861
Project: Camel
Issue Type: Improvement
Components: camel-langchain4j-agent, camel-langchain4j-embeddings,
camel-langchain4j-chat, camel-langchain4j-tools
Reporter: Claus Ibsen
The camel-langchain4j components currently have no OpenTelemetry integration
for AI/LLM-specific observability. Token usage, model name, finish reason, and
latency are not captured as span attributes.
Spring AI exposes this data through OpenTelemetry spans following the
OpenTelemetry Semantic Conventions for GenAI
(https://opentelemetry.io/docs/specs/semconv/gen-ai/), which enables tools like
Boot UI to build AI Usage dashboards showing token consumption, costs, model
distribution, and conversation tracking.
Camel should follow the same approach. When camel-opentelemetry (or
camel-opentelemetry2) is active, the langchain4j producers should enrich spans
with standard GenAI semantic convention attributes such as:
- gen_ai.system (e.g. openai, anthropic)
- gen_ai.request.model
- gen_ai.response.model
- gen_ai.usage.input_tokens
- gen_ai.usage.output_tokens
- gen_ai.response.finish_reasons
This would allow standard OpenTelemetry tooling (Jaeger, Grafana, etc.) and
Camel TUI to visualize AI/LLM usage across routes.
See also CAMEL-23860 which adds token data as exchange headers - this ticket is
about also propagating that data into OTel spans for observability tooling.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
