AI Ops Tech Leader: Agentic ML Reliability

Cisco; San Jose (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Cisco is seeking an AI Operations Engineering Technical Leader to drive the operationalization of complex autonomous agent architectures into secure, scalable production environments. You will define robust MLOps practices, pipelines, and model serving architectures to bring Agentic AI concepts into customer-facing products.

Design telemetry to monitor token usage, control compute costs, and ensure secure model performance.

Qualifications

• Experience deploying and operating large language models (OpenAI/Anthropic/Llama or similar).
• Strong background in software engineering and DevOps for scalable infrastructure.
• Proficient in Python and/or Java/J2EE; capable of API development with FastAPI/Flask/Spring.
• Hands-on with ML/CI/CD tooling and ML model tracking platforms.
• Familiarity with containerization (Docker) and orchestration (Kubernetes).

Responsibilities

• Drive operationalization of autonomous agent architectures into secure, scalable production environments.
• Define robust MLOps practices, pipelines, and model serving architectures for Agentic AI.
• Design telemetry to monitor token usage, compute costs, and model performance securely.
• Resolve infrastructure challenges on GPU clusters and containerized model inference.
• Mentor peers, guide architecture decisions, and ensure reliability and scalability.

Skills

Software Engineering & DevOps

LLM / Agentic AI experience

Python / Java / J2EE

API development (FastAPI/Flask/Spring)

CI/CD for ML / MLOps

Docker & Kubernetes

Education

Bachelor's degree in CS or related

Master's degree (preferred)

Tools

LangChain

LangSmith

Docker

Kubernetes

MLflow

Weights & Biases

ClearML

Job description

Cisco is seeking an AI Operations Engineering Technical Leader to drive the operationalization of complex autonomous agent architectures into secure, scalable production environments. You will define robust MLOps practices, pipelines, and model serving architectures to bring Agentic AI concepts into customer-facing products.

Design telemetry to monitor token usage, control compute costs, and ensure secure model performance.