Intermediate Agentic AI Engineer

CodeRoad; United States; Remote; USD 140,000 - 200,000; Full time

Benefits offered by this job

Remote work

Holidays off matching local calendar

Paid Time Off (PTO)

Health insurance assistance

Competitive USD compensation

Growth opportunities

Job summary

CodeRoad is seeking an Intermediate Agentic AI Engineer to design and scale a production-grade multi-agent platform. You will work primarily with Python and LangGraph to build a hierarchical agentic layer and implement dynamic context loading to optimize latency and token economics.

You will integrate tool-calling interfaces via MCP, strengthen security with OWASP Top 10, and improve observability with LangFuse while collaborating on automated evaluation harnesses and regression tests against

Qualifications

• 5+ years of professional software development with Python.
• 2+ years building production AI agents or complex LLM workflows with LangGraph or LangChain.
• Experience with modern foundation models (Anthropic Claude, OpenAI, Google Gemini) and prompt engineering.
• Hands-on with LLM observability platforms (LangFuse) and automated evaluation.
• Familiarity with containerized cloud environments (Azure, Docker).
• Experience with Microsoft 365 Agents SDK.
• Advanced English (written and spoken) mandatory.
• Strong ownership, problem-solving, and collaborative mindset.

Responsibilities

• Architect and deploy a hierarchical multi-agent workflow using LangGraph, transitioning legacy single-prompt implementations into modular agentic systems.
• Build domain-specific sub-agents featuring dynamic, on-demand context loading to optimize latency and token economics.
• Design and integrate scalable tool-calling capabilities and standardized connectors using the Model Context Protocol (MCP).
• Optimize LLM observability, reasoning traces, and cost-tracking systems with LangFuse.
• Lead automated evaluation harnesses and regression testing suites using PromptFlow or PromptFoo against benchmark datasets.
• Collaborate on implementing strict security guardrails and access controls following OWASP Top 10 standards for LLM applications.

Skills

Python

LangGraph

LangChain

Advanced English

Tools

LangFuse

Azure

Docker

Microsoft 365 Agents SDK

Job description

About CodeRoad

CodeRoad provides end-to-end software development services, helping businesses scale with ideal infrastructure solutions. From staff augmentation to dedicated IT teams and general software engineering, our nearshore technology services empower businesses to thrive in an ever-evolving digital landscape.

About the Role

As an Intermediate Agentic AI Engineer, you will serve as the technical backbone for building and scaling our production-grade multi-agent platform. Working primarily with Python and modern orchestration frameworks like LangGraph , you will design a hierarchical multi-agent layer, implement dynamic context-loading mechanisms, and integrate standardized tool-calling interfaces to transition legacy workflows into fully autonomous systems.

This role is critical to optimizing our AI ecosystem’s performance, token economics, and operational security. By establishing automated evaluation harnesses, robust observability stacks, and enterprise-grade security guardrails, your work will directly drive the reliability, safety, and scalable impact of high-performing AI agents across our organization.

Key Responsibilities

• Architect and deploy a hierarchical multi-agent workflow using LangGraph , successfully transitioning legacy single-prompt implementations into modular agentic systems.

• Build domain-specific sub-agents featuring dynamic, on-demand context loading to significantly optimize latency and token economics.

• Design and integrate scalable tool-calling capabilities and standardized connectors using the Model Context Protocol (MCP).

• Optimize LLM observability, reasoning traces, and cost-tracking systems by setting up and managing LangFuse.

• Lead the establishment of automated evaluation harnesses and regression testing suites using tools like PromptFlow or PromptFoo against benchmark datasets.

• Collaborate on implementing strict security guardrails and access controls following OWASP Top 10 standards for LLM applications.

Requirements

• 5+ years of professional software development experience, with a primary focus on Python.

• 2+ years of hands‑on experience building production AI agents or complex LLM workflows using LangGraph or LangChain.

• Tech Stack: Strong practical experience with modern foundation models ( Anthropic Claude , OpenAI , Google Gemini ), open‑weights models, and advanced prompt engineering techniques.

• Observability & Eval: Hands‑on experience with LLM observability platforms (e.g., LangFuse ) and automated evaluation frameworks.

• Infrastructure: Familiarity with containerized cloud environments including Azure and Docker.

• Ecosystem: Hands‑on experience with the Microsoft 365 Agents SDK.

• Soft Skills: High ownership mindset, strong problem‑solving initiative, and an empathetic, collaborative team approach.

• Language: Advanced English (written and spoken) is mandatory.

Nice to Have

• Experience working with Retrieval‑Augmented Generation ( RAG ) pipelines and vector database integrations (e.g., Pinecone , Weaviate , Qdrant ).

• Familiarity with enterprise AI security frameworks, data privacy compliance, and prompt injection mitigation strategies.

• Exposure to serverless architectures and microservices deployments on cloud platforms.

What You’ll Love

• 100% Remote work environment.

• Holidays off matching local calendar standards.

• Generous Paid Time Off (PTO).

• Health insurance assistance.

• Competitive USD compensation.

• Clear growth opportunities and continuous learning support.