Generative AI Applications Engineer (Agents & RAG)

Role overview A hands-on engineering role building secure, production-grade generative AI applications for federal sector programs. The work spans agentic workflows, retrieval-augmented generation, prompt and policy design, LLM evaluation, and integration with cloud AI platforms. It is delivery-focused: teams ship features in weeks and operate them with the same rigor as any other production service.

Responsibilities - Design and ship mission-grade GenAI features, including agentic workflows and RAG systems that target low hallucination, tight p95 latency, and predictable cost. - Apply patterns from LangChain, LlamaIndex, Semantic Kernel, and similar frameworks to handle task decomposition, tool use, guardrails, and recovery or fallback strategies. - Integrate with cloud AI services such as AWS Bedrock, Azure OpenAI, Google Vertex AI, and Amazon Kendra; leverage managed services like Document AI, Gemini, and Gemma. - Build RAG pipelines with vector databases such as Pinecone, Weaviate, OpenSearch, pgvector, FAISS, or Chroma, including chunking, metadata design, and retrieval-quality evaluations. - Define SLIs and SLOs across quality, latency, safety, and cost; run on-call rotations, postmortems, and FinOps optimization. - Ship reusable platform components, including SDKs, CI/CD templates, Terraform modules, and evaluation harnesses that accelerate multiple mission teams. - Deliver into hybrid, restricted, or air-gapped environments using Zero Trust principles and audit-ready controls.

Requirements - Demonstrated experience building and deploying a production GenAI application such as a chatbot, copilot, or enterprise assistant. - Hands-on work with large language models including GPT, Claude, Llama, Gemini, or Mistral, accessed via APIs or self-hosted. - Experience designing RAG solutions with embeddings, vector databases, and semantic search for enterprise or mission data. - Familiarity with AI application frameworks such as LangChain, LangGraph, LlamaIndex, Semantic Kernel, or DSPy. - Strong Python development skills for building and integrating AI/ML applications. - Track record of owning AI solutions through the full production lifecycle, including deployment and operational support. - U.S. citizenship.

Nice to have - Integration experience with leading cloud AI services or on-premises inference stacks. - Background in LLM evaluation, prompt authoring and testing, A/B experimentation, and applied ML.

Benefits and work setup The role supports confidential federal programs, so program specifics are not disclosed publicly and are shared later in the process. The team emphasizes hands-on growth, certifications, and industry training, and measures success by latency, reliability, safety, and cost.