Agentic Engineer

About the role

The Agentic Engineer will be responsible for designing, building, and deploying production-grade agentic AI systems that autonomously execute multi-step workflows across our insurance operations. You will work at the intersection of LLM orchestration, tool integration, and enterprise systems to bring agent-native capabilities to life. Responsibilities span the full agentic stack: from defining agent anatomy (triggers, loops, tools, memory, human gates, and observability) to hardening agents for reliability, safety, and governance in a regulated environment.

Responsibilities

• Design and build end-to-end agentic workflows that autonomously execute multi-step business processes, including FNOL processing, claims triage, and audit automation.
• Develop and maintain agent orchestration patterns using frameworks such as LangChain, LangGraph, or custom-built agent loops, integrating LLMs with enterprise tools, APIs, and data platforms.
• Implement robust tool use layers, including function calling, MCP integrations, and structured output validation, to connect agents to internal systems and external services.
• Design and enforce human-in-the-loop gate patterns, escalation logic, and approval checkpoints to ensure agents operate within defined authority boundaries in a regulated insurance environment.
• Build and manage memory tier strategies for agents, including in-context, external retrieval (RAG/vector stores), and episodic memory patterns appropriate to each workflow.
• Instrument agents with full observability: structured logging of agent reasoning traces, tool call outcomes, token usage, latency, and failure modes to support monitoring, debugging, and audit requirements.
• Implement prompt injection defenses, output validation, and agent identity controls to maintain security and prevent misuse or unintended agent behavior in production.
• Evaluate and benchmark agent performance through end-to-end testing, red-teaming, and simulation of edge cases to validate reliability before production deployment.
• Partner with Product Owners and Enterprise Architects to translate business workflows into agent-native designs, ensuring alignment on scope, authority, and fallback behaviors.
• Champion CI/CD best practices for agent deployments, including versioned prompts, tool schemas, and agent configurations with rollback capability across Dev, QA, and Prod environments.
• Contribute to and extend internal agentic infrastructure, including shared tool registries, reusable agent templates, and platform abstractions that accelerate agent development across the team.
• Other duties as assigned.

Requirements

• Previous experience building and deploying AI agents, LLM-powered applications, or agentic workflow systems in a production-like environment is required.
• Demonstrated experience with agentic orchestration frameworks (e.g., LangChain, LangGraph, AutoGen, CrewAI, or custom agent loops) and LLM tool/function calling patterns.
• Strong understanding of machine learning techniques, software architecture and distributed systems, including how agents interact with APIs, databases, message queues, and external services at runtime.
• Hands-on experience with RAG architectures, vector databases (e.g., Pinecone, pgvector, Chroma), and embedding strategies for grounding agent responses in enterprise knowledge.
• Familiarity with AI safety and trust concepts relevant to agentic systems: prompt injection, output sanitization, privilege minimization, and human oversight patterns.
• Ability to write robust, production-quality code in Python; proficiency with async patterns and REST APIs for building reliable agent tool integrations.
• Experience deploying and serving open-source LLMs using frameworks such as vLLM, llama.cpp, TensorRT-LLM, or similar inference engines.
• Familiarity with model optimization techniques (quantization, batching strategies, GPU memory management) for production workloads.
• Experience deploying containerized AI services using Docker and Kubernetes (or OpenShift/EKS) within cloud environments (AWS/Azure), including managing inference infrastructure and API gateways.
• Exemplary written, verbal, listening, and interpersonal communication skills.
• Excellent analytical, problem-solving, and decision-making skills.
• Ability to work both independently and collaboratively as part of a team.

Work Environment

• Remote: This role is a remote (work from home (WFH)) opportunity, and only open to candidates currently located in the United States and able to work without sponsorship. It requires a suitable space that provides a private and quiet workplace.
• Expected Work Hours: Schedules are set to accommodate the requirements of the position and the needs of the organization and may be adjusted as needed.
• Travel: May be required to travel to off-site location(s) to attend meetings, as necessary.

Pay & Benefits

• Salary Range: $105,000 - $160,000 and a comprehensive benefits package.