AI Infrastructure Architect

Project Role

Experience and Education

• The role requires 5–8 years of software or platform engineering experience.
• At least 2 years of experience focused on AI, machine learning, or agentic systems is required.
• The role requires 15 years of full-time education.

Work Mode

• The source states the work mode as remote or hybrid, without selecting one definitive arrangement.

Key Responsibilities

• Build multi-agent orchestration systems using Vertex AI, Agent Builder, LangGraph, and AutoGen.
• Design resilient pipelines for LLM reasoning, tool calling, memory, and human-in-the-loop flows.
• Lead RAG systems using Vertex AI Search, Matching Engine, and enterprise vector stores.
• Define secure GCP infrastructure standards across GKE, Cloud Run, Eventarc, and Apigee.
• Drive MLOps for evaluation, observability, tracing, logging, and improvement loops.
• Translate business requirements into agentic solutions with product, data science, and platform teams.
• Conduct code and architecture reviews.
• Mentor developers with 3–5 years of experience.
• Contribute to internal developer platforms, SDKs, and reusable agent components.

Required Skills and Qualifications

• Deep expertise in Vertex AI, Gemini, Vertex AI Pipelines, Feature Store, GKE, and Pub/Sub.
• Production-grade LLM application experience with tool use, multi-turn reasoning, and agent memory.
• Proficiency in Python and strong understanding of asynchronous patterns, concurrency, and distributed systems.
• Hands-on experience with ADK, LangChain, LangGraph, CrewAI, or comparable agentic frameworks.
• Experience with vector databases, embedding pipelines, and hybrid search architectures.
• Strong GCP security knowledge covering IAM, Workload Identity, VPC Service Controls, and Secret Manager.
• Experience delivering high-availability and low-latency services with SLO or SLA ownership.

Must-Have Skills

• Agent development is listed as the must-have skill.