LR
Principal Engineer Generative & Agentic AI
Accepting applicationsLumen Resources · United States
Full-Time Mid_senior AIMachine Learningaiate
Posted
18 May
Category
Manufacturing
Experience
Mid_senior
Country
United States
We are partnering with a high-growth, mission-driven company expanding rapidly in the US market to find their first Principal AI Engineer. This is a greenfield opportunity where you will be the technical architect and builder of a production-grade multi-agent AI platform from the ground up, reporting directly to the VP of AI.
This is not a research role. This is not an orchestration workflow role. This is a hands-on, production-first engineering role for someone who has built true multi-agent systems, shipped them to production, watched them fail, fixed them, and scaled them across real enterprise workflows.
Must Have:
Proven experience building and deploying true multi-agent systems in production systems where agents have distinct roles, reason independently, and coordinate with each other, not a single orchestrator calling tools in sequence
Deep hands-on experience with agent orchestration frameworks, LangGraph, AutoGen, CrewAI, Amazon Bedrock Agents, Google ADK, or equivalent
Real production experience with state management and memory architecture, short-term session memory, long-term persistent memory, state passing between agents, and recovery from mid-workflow failures
Ability to describe specific production failures, what broke, what the system did, what you changed architecturally, and what the measurable improvement was
Experience designing evaluation frameworks, LLM-as-a-judge, confidence thresholds, inline evaluation at intermediate outputs, not just end-of-pipeline monitoring
Demonstrated platform ownership, systems that other teams or workflows have built on top of, not just purpose-built point solutions
Strong Preference:
MS or PhD in Computer Science, Machine Learning, or a related technical field from a recognised institution
Experience with MLOps, LLMOps, CI/CD with deployment gating for AI systems
Hands-on experience with AWS Bedrock, Azure AI Foundry, or GCP Vertex AI in production
Show more Show less
This is not a research role. This is not an orchestration workflow role. This is a hands-on, production-first engineering role for someone who has built true multi-agent systems, shipped them to production, watched them fail, fixed them, and scaled them across real enterprise workflows.
Must Have:
Proven experience building and deploying true multi-agent systems in production systems where agents have distinct roles, reason independently, and coordinate with each other, not a single orchestrator calling tools in sequence
Deep hands-on experience with agent orchestration frameworks, LangGraph, AutoGen, CrewAI, Amazon Bedrock Agents, Google ADK, or equivalent
Real production experience with state management and memory architecture, short-term session memory, long-term persistent memory, state passing between agents, and recovery from mid-workflow failures
Ability to describe specific production failures, what broke, what the system did, what you changed architecturally, and what the measurable improvement was
Experience designing evaluation frameworks, LLM-as-a-judge, confidence thresholds, inline evaluation at intermediate outputs, not just end-of-pipeline monitoring
Demonstrated platform ownership, systems that other teams or workflows have built on top of, not just purpose-built point solutions
Strong Preference:
MS or PhD in Computer Science, Machine Learning, or a related technical field from a recognised institution
Experience with MLOps, LLMOps, CI/CD with deployment gating for AI systems
Hands-on experience with AWS Bedrock, Azure AI Foundry, or GCP Vertex AI in production
Show more Show less