Job Description

Job Summary

Senior subject-matter expert for the operations, observability, and lifecycle management of AI agents in production (“AgentOps” / LLMOps). Owns the frameworks and practices to safely deploy, monitor, evaluate, and continuously improve live agents — ensuring reliability, safety, cost-efficiency, and business-KPI performance across the Intel Agent Factory.

Key Responsibilities

• Define and operate the AgenticOps framework: agent registry, versioning, guarded rollout, and rollback for production agents.
• Establish continuous evaluation and monitoring: quality, autonomy, safety (guardrails, Model Armor), latency, cost, and reuse metrics.
• Implement observability and tracing for multi-agent systems (Agent Engine Observability, Cloud Monitoring/Logging/Trace).
• Own the 5-gate validation-to-production process and post-release escape management for delivered agents.
• Design human-in-the-loop (HITL) supervision, feedback loops, and automated pre-production simulations for safe rollout.
• Track and report agent business KPIs (CSAT, TAT, MTTR, cost savings) via AgentScore / Agent 360 dashboards.
• Drive cost governance for agent runtimes: model tiering, context caching, batch/flex inference, budget caps and alerts.
• Collaborate with DevOps SME (deploy) and AI & Data SME (grounding) to close the build-deploy-operate-improve loop; advise Intel on AgenticOps ownership transfer.

Skill Requirements

• Strong LLMOps / MLOps / AgentOps experience operating GenAI or agentic systems in production.
• Hands-on with Google Cloud agent runtimes: Vertex AI, Agent Engine, and observability tooling.
• Agent evaluation and safety: eval frameworks, guardrails, Model Armor, HITL, prompt/robustness testing.
• Monitoring, tracing, and reliability engineering (SRE) for AI workloads.
• Cost governance and performance tuning for LLM/agent workloads.
• Proficiency in Python; strong grasp of agent lifecycle and governance.

Other Requirements

• Experience with ADK, A2A, MCP, and multi-agent orchestration in production.
• BigQuery/Looker for agent analytics and KPI dashboards.
• Responsible-AI, model governance, and audit/compliance frameworks.
• Prior enterprise-scale AI platform operations experience.
• 9–12+ years in ML/AI platform operations, SRE, or LLMOps with production agentic/GenAI exposure (Tier 5–6).
• Google Cloud Professional (ML/DevOps) certification preferred.
• Offshore (India) with US overlap, or Onsite (USA) as required.