Production GenAI Engineer: Agents & RAG

Accenture Federal Services; Washington (District of Columbia); On-site; USD 103,200 - 203,400; Full time

Benefits offered by this job

Collaborative work environment

Hands-on growth opportunities

Certifications and industry training

Job summary

A leading technology firm in federal services is looking for a hands-on role to design and deploy GenAI applications. The ideal candidate will have strong Python skills, experience with LLMs, and end-to-end ownership of production systems. Responsibilities include building workflows tailored to mission data, evaluating AI models, and enhancing production quality. A commitment to U.S. Citizenship is essential. This position offers competitive compensation within a supportive team environment.

Qualifications

• End-to-end ownership of production systems: integration, deployment, observability, incident response.
• Hands-on experience with LLMs, transformer-based apps, and RAG in production.
• Experience with vector search and retrieval.

Responsibilities

• Design & ship mission grade GenAI applications.
• Apply patterns from LangChain/LlamaIndex/Semantic Kernel.
• Implement with AWS Bedrock, Azure OpenAI, Google Vertex AI.
• Build retrieval pipelines and vector search.
• Instrument metrics/logs and run A/B experiments.

Skills

End-to-end ownership of production systems

Hands-on experience with LLMs

Strong Python

Experience with vector search

U.S. Citizenship

Tools

AWS Bedrock

Azure OpenAI

Docker

Kubernetes

Linux Bash

Job description

A leading technology firm in federal services is looking for a hands-on role to design and deploy GenAI applications. The ideal candidate will have strong Python skills, experience with LLMs, and end-to-end ownership of production systems. Responsibilities include building workflows tailored to mission data, evaluating AI models, and enhancing production quality. A commitment to U.S. Citizenship is essential. This position offers competitive compensation within a supportive team environment.