Bedrock Support Engineer

ASCENDING

1d • Onsite

About The Position

We are seeking a Support Engineer to manage and scale our Amazon Bedrock infrastructure. This role is dedicated to the operational excellence of Foundation Models (FMs), focusing on the automation of model deployments, optimization of RAG (Retrieval-Augmented Generation) pipelines, and maintaining high-performance AI workflows. You will ensure that our Generative AI environment is secure, cost-effective, and seamlessly integrated into our CI/CD ecosystem.

Requirements

• Experience: Experience in DevOps or Cloud Engineering, with a significant recent focus on AI/ML infrastructure support.
• Bedrock Expertise: Hands-on experience configuring Amazon Bedrock (Knowledge Bases, Agents, and Guardrails).
• DevOps Tools: Expert proficiency in Terraform, Jenkins/GitLab CI, and Python (Boto3).
• AWS AI Services: Deep technical knowledge of supporting services, including AWS Lambda, Amazon OpenSearch, and S3.
• Security: Strong understanding of IAM granular permissions specifically for AI service access and data privacy.
• Location: Ability to work onsite 5 days a week in McLean, VA.

Nice To Haves

• Azure OpenAI Services

Responsibilities

• Amazon Bedrock Operations
• Model Provisioning: Manage Amazon Bedrock model access and configure Provisioned Throughput to ensure consistent performance for production LLM workloads.
• RAG Management: Maintain and optimize Knowledge Bases for Amazon Bedrock, including the data ingestion pipelines and vector store integrations (e.g., OpenSearch Serverless).
• Agent Orchestration: Support and troubleshoot Agents for Amazon Bedrock, ensuring Lambda-based action groups and API integrations are functioning correctly.
• Governance & Safety: Implement and monitor Guardrails for Amazon Bedrock to filter sensitive content and enforce responsible AI policies across all applications.
• GenAI Infrastructure & Automation
• AI-Specific IaC: Develop and maintain Terraform or AWS CDK modules specifically for Bedrock components, IAM roles, and associated data resources (S3, Lambda).
• CI/CD for LLMs: Build automated pipelines for model testing, evaluation (using Model Evaluation on Amazon Bedrock), and seamless promotion of AI-enabled features.
• Private AI Networking: Configure and troubleshoot VPC Interface Endpoints (PrivateLink) for Bedrock to ensure AI data traffic remains within the private network.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed