AgentOps Engineer jobs
AgentOps Engineers keep agents healthy once they're live. They build the monitoring, tracing, eval pipelines, and guardrails that catch regressions, control cost and latency, and keep behavior within bounds. Think SRE/MLOps discipline applied to non-deterministic, tool-using systems.
AgentOps Engineer · Agent Operations Engineer · LLMOps Engineer · ML Platform Engineer (Agents)
Observability & tracing · Eval pipelines · Cost & latency optimization · CI/CD for prompts · Incident response
What a AgentOps Engineer does
- •Build tracing and observability for tool-using, non-deterministic agents
- •Stand up automated eval pipelines and wire eval gating into CI/CD
- •Own cost and latency budgets, dashboards, and alerting for agent fleets
- •Run incident response and write runbooks for agent failures and regressions
- •Enforce guardrails, rate limits, and rollout/rollback for prompt and model changes
Leveling
- Mid
- Operates tracing/eval tooling for a product; responds to incidents.
- Senior
- Owns the observability + eval platform and cost/latency posture.
- Staff+
- Defines AgentOps standards and tooling across the org.
Salary by level
Full salary breakdown →US total-cash ranges synthesized from aggregated public job postings and market data as of 2026. Individual offers vary widely by location, company stage, and equity; treat these as directional bands, not guarantees.
Open AgentOps Engineer roles (44)
Search & filter →Customer Success - EMEA
ElevenLabs
Forward Deployed Engineer (m/f/d)
Blueocean Technologies Gmbh
Backend QA Engineer
Jobgether
Werkstudent:in Logistik (m/w/d)
PURELEI GmbH
Werkstudent HSEQ (m/w/d)
Vulcan Energie Ressourcen GmbH
AgentOps Engineer FAQ
- What is AgentOps?
- AgentOps is the operations discipline for running AI agents in production — monitoring, tracing, eval pipelines, cost/latency control, and guardrails. It's SRE/MLOps adapted to non-deterministic, tool-using systems.
- How is AgentOps different from MLOps?
- MLOps focuses on training, deploying, and monitoring models. AgentOps adds the agent layer: evaluating multi-step tool-using behavior, tracing decisions, controlling per-run cost/latency, and guarding against unsafe actions.
- How much does an AgentOps engineer make?
- In the US, total-cash ranges run roughly $150k–$200k (mid), $185k–$250k (senior), and $240k–$330k+ (staff and above).
Learn more
- AgentOps, Explained: The SRE for AI Agents (2026)
What AgentOps is, how it differs from MLOps, the core responsibilities and stack, plus 2026 US salary bands for AgentOps engineers.