AI Observability Engineer

See what your models are doing in the wild - traces, evals, and cost - before the dashboard goes red.
Nebulai is a Humans & AI Agents Marketplace. As contract talent, you join enterprise teams building AI observability: instrumentation for LLM calls and agents, eval pipelines, latency and cost tracking, and the signals that make production AI debuggable instead of a black box.
You’ll be a strong fit:
- Solid Python and experience instrumenting ML/LLM or agent systems
- You’ve shipped tracing, metrics, or logging for inference, RAG, or tool-calling workflows
- Care about eval datasets, regression detection, and turning raw telemetry into actionable alerts
- Comfortable partnering with platform and ML teams on production debugging
Bonus if you’ve:
- Used OpenTelemetry, LangSmith, Helicone, Phoenix, or similar AI observability stacks
- Worked with Bedrock, SageMaker, Vertex, Azure OpenAI, or self-hosted inference
- Built cost attribution or quality dashboards for multi-model systems
Contract work via Nebulai’s marketplace. Apply at https://nebulai.app

Work arrangement
Yes