Lead Agentic AI Engineer: Multi-Agent Systems

Ethereum Technologies

Irvine (CA)

On-site

USD 180,000 - 240,000

Full time

7 hours ago

Be an early applicant

Get a reply from this recruiter

Your resume and cover letter, tailored to exactly what they’re hiring for.

Get past ATS filters

Job summary

Ethereum Technologies is seeking an experienced AI engineer to design and implement enterprise-grade agents and interoperable multi-agent workflows in Irvine, CA. You will translate business use cases into reasoning agents, manage memory and context, and integrate with trusted tools and REST APIs within secure, scalable pipelines.

The role emphasizes A2A collaboration, tooling orchestration, and production readiness, with exposure to Gemini models and RAG techniques.

Qualifications

Strong Python development skills and hands-on experience with LLM or Generative AI applications.
Experience with agentic patterns: reasoning, orchestration, tool calling and workflow execution.
Experience with Google ADK, Vertex AI, Agent Engine and Gemini, or comparable agent frameworks with ability to transition to this stack.
Understanding of Agent-to-Agent (A2A) architecture and multi-agent patterns, including agent discovery, capability exchange, task lifecycle, asynchronous interactions and secure context handoff.
Understanding of RAG, embeddings, vector search, chunking, grounding and retrieval evaluation.
Understanding of context and memory management for conversational and multi-agent applications.
Experience consuming REST APIs and integrating external services; familiarity with MCP for agent-to-tool connectivity.
Working knowledge of GCP, Git/GitHub, Docker or containers, CI/CD and automated testing.

Responsibilities

Design and build enterprise-grade AI agents and interoperable multi-agent workflows using modern agentic AI frameworks. Translate defined business use cases into agents that can reason, use governed enterprise knowledge, maintain context, invoke approved tools and APIs, and collaborate with other agents through Agent-to-Agent (A2A) patterns. Contribute to rapid pilots and production-ready implementations using reusable platform capabilities.
Design and develop enterprise AI agents based on defined business workflows and acceptance criteria.
Build agent workflows using Google ADK and Vertex AI / Agent Engine.
Use Gemini models for reasoning, planning, summarization and structured response generation.
Develop prompts, prompt templates and structured outputs; support prompt testing and evaluation.
Implement RAG capabilities using approved enterprise knowledge sources, chunking, embeddings and retrieval.
Implement session context, short-term memory and long-term memory patterns where required.
Integrate agents with approved MCP tools, REST APIs and external AI services.
Design agent-to-agent collaboration using A2A patterns, including agent discovery, capability definition, task delegation, context exchange, status handling and response coordination.
Implement orchestration, tool selection, multi-step execution and human-in-the-loop checkpoints across single-agent and multi-agent workflows.
Apply input/output validation, guardrails and secure handling of sensitive information.
Build modular, reusable and testable Python components; support deployment and operational troubleshooting.
Hands-on exposure to implementing A2A-compatible agents or interoperable multi-agent workflows.
Exposure to BigQuery, AlloyDB, Vertex AI Search, vector databases or knowledge graphs.
Familiarity with Model Armor, AI guardrails, Responsible AI controls, IAM, RBAC or ABAC.
Experience taking an AI proof of concept through pilot deployment and production readiness.