AI Operations Engineer

About the Team

Join Cisco's Commerce Intelligence Data & Analytics team. We deliver seamless intelligence, advanced analytics, and cutting-edge agentic experiences to drive high-impact business decisions and automated actions. Our mission is building a scalable, unified, AI-ready data foundation.

Your Impact as an AI Operations Engineer

You will design, build, and operate autonomous agents powering Cisco's commerce operations by translating operational requirements into reliable multi-step workflows orchestrating LLM reasoning, tool calling, retrieval, and platform integrations. You own the full lifecycle from agent architecture to production monitoring.

Key Responsibilities

• Architect production AI agents (multi-agent/human-in-the-loop) automating commerce ops using frameworks like LangGraph, AutoGen, or CrewAI.
• Build Agent-to-Agent (A2A) integrations and interoperability protocols for reliable delegation and handoffs.
• Design robust API integrations (REST/GraphQL), event-driven services, authentication, rate limiting, and resilient error handling connecting agents to enterprise data sources.
• Manage end-to-end lifecycle of LLM apps using Snowflake Cortex including model selection, fine-tuning, deployment, versioning, and decommissioning.
• Architect agentic workflows with Snowflake Native Apps for sophisticated tool-calling and state management.
• Build RAG pipelines over documents and structured data; maintain low-latency vector databases and semantic search architectures.
• Create automated monitoring for LLM drift, hallucination rates, and token usage triggering self-healing alerts or retraining workflows.
• Optimize Snowflake credit consumption via auto-scaling, query profiling, and multi-cluster strategies ensuring cost-efficiency at scale.
• Implement guardrails, human-in-the-loop gates, and LLM routing/failover for reliability and safety under real operational conditions using advanced prompt engineering.
• Own the AI/ML CI/CD pipeline with rigorous version control, automated testing, and progressive rollout strategies across dev/staging/prod environments.