Skip to content

engineering · 1 min read

OpenTelemetry GenAI Semantic Conventions in Practice

OTel's GenAI semantic conventions standardise AI observability. Here is what they cover and how Bhogar AI implements them.

BABhogar AI TeamProduct & Engineering

OpenTelemetry now has stable semantic conventions for GenAI. Adopting them future-proofs your observability and gives you portability across vendors.

Why it matters

The conventions cover prompt, response, model, tokens, errors, refusals - the things you actually want to filter and aggregate on.

How Bhogar AI approaches it

Bhogar AI emits OTel-compliant traces and metrics with the GenAI conventions out of the box, exportable to any OTLP backend.

  • Prompt and response captured as span attributes
  • Token counts and cost as metrics
  • Refusal and error categorisation
  • OTLP export to any backend
  • Sampling tuned for AI workloads

What you get

AI observability that integrates with the wider OTel ecosystem and survives vendor changes.

See Bhogar on your own data

Book a 45-minute working session. We connect one of your sources, build one agent, run one governed workflow, and review the trace together.