Generative AI Applications Engineer (Agents & RAG)

Role overview Build and operate production generative AI applications for sensitive public-sector use cases. The work spans agent workflows, retrieval-augmented generation (RAG), prompt and policy design, model evaluation, and deployment; success is measured through quality, safety, reliability, latency, and cost. Model training is not part of the role.

Responsibilities - Assess use cases and deliver agentic workflows and RAG applications, with controls to reduce inaccurate responses and manage latency and cost. - Orchestrate tools and task flows using agent frameworks; design guardrails, recovery paths, and fallbacks. - Integrate managed AI platforms and services, and compare models for quality, safety, speed, and cost. - Build retrieval pipelines, including data preparation, chunking, metadata, vector search, and retrieval evaluation. - Instrument applications with metrics, logs, and traces; run experiments and maintain incident response practices. - Define service objectives for quality, safety, latency, and cost; improve operations and develop reusable SDKs, deployment templates, and evaluation tools.

Requirements - Active U.S. citizenship. - Experience building and deploying a production generative AI application and supporting it through its lifecycle. - Strong Python development skills for AI application work. - Hands-on experience with LLMs through APIs or self-hosted deployments. - Experience implementing RAG with embeddings, vector databases, and semantic search. - Familiarity with agent frameworks such as LangChain, LangGraph, LlamaIndex, Semantic Kernel, or similar.

Nice to have Experience with cloud AI services or on-premises inference, model evaluation, prompt testing, and A/B experimentation.

Benefits and work setup The role may involve hybrid, restricted, or air-gapped environments, with Zero Trust principles and audit-ready controls. The listing describes access to training and certifications but does not provide specific benefit details.