RAG Application Engineer

About the Role

As a RAG Application Engineer, you will build the application layer for Retrieval-Augmented Generation (RAG) systems. You will implement the end-to-end query workflow, including embedding generation, hybrid retrieval, reranking, citation assembly, and request-level guardrails. You will design and maintain RAG runtime logic in a secure, cloud-based environment, develop API endpoints, manage model integrations, enforce authentication and authorization controls, and implement robust retry, timeout, and error-handling strategies. You will containerize services and partner with Cloud, Security, Data, and Site Reliability Engineering teammates.

Key Responsibilities

• Implement the end-to-end query workflow for RAG systems.
• Build the application layer to integrate search, models, and metadata.
• Develop API endpoints.
• Manage model integrations.
• Enforce authentication and authorization controls.
• Implement robust retry, timeout, and error-handling strategies.
• Containerize services.
• Partner with Cloud, Security, Data, and Site Reliability Engineering teammates.

Requirements & Tech Stack

• Experience building back-end or platform services using programming languages such as Python or Java.
• Experience designing or integrating RAG workflows, including embeddings, hybrid retrieval, vector search, and reranking.
• Experience implementing API request pipelines, orchestration logic, and service routing for AI model and search calls.
• Experience enforcing authentication and authorization within service logic.
• Experience developing resilient services with retry, backoff, timeout, and error-handling strategies.
• Experience containerizing applications using Docker.
• Experience deploying to ECS, EKS, or Kubernetes.
• Experience building telemetry, including structured logs, metrics, and traces.
• Ability to translate RAG requirements into clean, maintainable application logic.
• TS/SCI clearance.
• HS diploma or GED.

Nice If You Have

• Experience integrating Amazon Bedrock model inference into application workflows.
• Experience implementing hybrid retrieval with OpenSearch, including BM25 and vector search.
• Experience developing citation assembly logic, grounding validation, or model guardrails.
• Experience designing API gateways, rate limiting, caching, or other high-availability patterns.
• Experience working with message queues or asynchronous workflows, including Kafka or Amazon SQS.
• Experience contributing to evaluation data pipelines or partnering with evaluation scientists.
• Experience using infrastructure-as-code frameworks such as Terraform, CloudFormation, or AWS CDK.
• Ability to collaborate across multiple engineering disciplines.
• Possession of excellent communication skills.