AI Engineer

Metas Solutions•, GA

About The Position

Metas Solutions is hiring an AI Engineer to build, enhance, and support production generative AI web applications, including retrieval-based knowledge assistants, chatbots, and agent workflows. The role involves implementing features, integrating enterprise data and services, troubleshooting issues, and improving application quality. The engineer will collaborate with senior AI engineers, architects, product owners, data engineers, and DevSecOps teams to deliver secure, maintainable production applications, and assess and suggest areas for improvement and potential solutions. The position requires developing Python backend services and AI integrations using Azure OpenAI, Azure AI Foundry, and related Azure services; implementing retrieval-augmented generation pipelines and agent workflows; integrating applications with APIs, enterprise data sources, databases, cloud services, and MCP-based tools; and evaluating application quality, reliability, latency, and cost. It also includes adding observability, troubleshooting defects, supporting deployments and production monitoring, participating in code reviews, documenting implementation decisions, and collaborating with engineering and DevSecOps teams. Applicants must be U.S. citizens and able to obtain a U.S. Government Public Trust 5 security clearance within a reasonable period. The role offers a market-competitive salary commensurate with experience and education. Metas Solutions provides a comprehensive benefits package and notes that use of AI or other third-party resources to generate or assist interview responses is not permitted and may result in disqualification.

Requirements

At least 3 years of technical experience, including developing and supporting production software applications.
At least 1 year of hands-on generative AI development, including RAG-based assistants or chatbots.
Experience with Python backend development, including asynchronous APIs, using FastAPI, Pydantic, or comparable tools.
Experience with Git, code reviews, automated testing, CI/CD, and Agile team delivery.
Experience with Azure, including Azure OpenAI, Azure AI Search, and integration with databases and external services.
Ability to demonstrate agent workflow experience with tool calling, structured outputs, and context management using PydanticAI, Microsoft Agent Framework, LangGraph, Foundry Agent Service, or similar tools.
Ability to conduct LLM evaluation and troubleshooting using test datasets, logs, and traces to assess quality, latency, and cost.
Hands-on RAG experience with document processing, chunking, embeddings, vector or hybrid search, and citations.
Bachelor's degree in computer science, engineering, or a related field.
Must be a U.S. citizen and able to obtain a U.S. Government security clearance (Public Trust 5) within a reasonable period.

Nice To Haves

Experience with Langfuse, Pydantic Logfire, or similar AI observability tools, and OpenTelemetry.
Experience with advanced RAG, multi-agent workflows, or MCP integrations.
Experience with Docker, Azure deployment services, or frontend development using React, Angular, or TypeScript.
Knowledge of open-source LLM hosting using vLLM, SGLang, or similar frameworks.
Knowledge of LLM fine-tuning and optimization, including PEFT, LoRA/QLoRA, and quantization.

Responsibilities

Build, enhance, and support production generative AI applications and web-based AI solutions.
Develop Python backend services and AI integrations using Azure OpenAI, Azure AI Foundry, and related Azure services.
Implement RAG pipelines, including document processing, chunking, embeddings, search, context preparation, and source citations.
Implement agent workflows using tool or function calling, structured outputs, conversation context, error handling, execution limits, and human review where needed.
Integrate AI applications with APIs, enterprise data sources, databases, cloud services, and MCP-based tools.
Test and evaluate retrieval relevance, answer correctness and grounding, reliability, latency, and cost.
Add logging, tracing, and metrics to monitor model calls, retrieval steps, tool execution, errors, and token usage.
Troubleshoot application issues, fix defects, and support deployments and production monitoring.
Participate in code reviews, document implementation decisions, and collaborate with engineering and DevSecOps teams.
Assess and suggest areas for improvement and potential solutions.