AI Engineering Leader: Agent Orchestration & LLM

BDO USA, LLP; Grand Rapids (MI); On-site; USD 130,000 - 160,000; Full time

Job summary

BDO USA, LLP is seeking an AI Engineering Manager to lead AI capabilities across the DT&I portfolio. You will design, build, deploy, and operate agent architectures and LLM-powered features within BDO's Azure ecosystem, ensuring accuracy, reliability, and cost-effectiveness.

You will align practices with governance, responsible-AI standards, and security policies while guiding cross-functional teams. Key areas include agent orchestration, RAG pipelines, and LLMOps across the Advantage product

Qualifications

• Seven or more years in software engineering or related tech fields.
• Three or more years building and deploying production AI/ML or LLM applications.
• Three or more years with AI orchestration frameworks (Semantic Kernel, LangChain) and Azure AI services.

Responsibilities

• Owns model selection, routing, fallback, and cost/latency optimization for AI features.
• Build evaluation harnesses, regression tests, and guardrails for accuracy and cost targets.
• Develop agent architectures and production-grade AI capabilities across the product stack.
• Design observability, error handling, and resilient AI workflows.
• Collaborate with cross-functional teams to deliver integrated AI solutions.

Skills

Semantic Kernel

Azure AI Foundry

C#/.NET

React

LangChain

RAG

Observability

Education

Bachelor's degree in CS/IS/Engineering

Tools

Docker

Git

Azure DevOps

Azure Container Apps

Azure AI Foundry

Job description

BDO USA, LLP is seeking an AI Engineering Manager to lead AI capabilities across the DT&I portfolio. You will design, build, deploy, and operate agent architectures and LLM-powered features within BDO's Azure ecosystem, ensuring accuracy, reliability, and cost-effectiveness.

You will align practices with governance, responsible-AI standards, and security policies while guiding cross-functional teams. Key areas include agent orchestration, RAG pipelines, and LLMOps across the Advantage product