Hybrid AI Engineer - Agentic AI on AWS Bedrock

ZealHire Inc.; Los Angeles (CA); Híbrido; USD 110 000 - 180 000; Tempo integral

Resumo da oferta

ZealHire Inc. is seeking an experienced AI Engineer to design, deploy, and monitor production-grade agents using Claude Agent SDK and AWS Bedrock AgentCore.

You will translate complex business needs into reliable, observable AI workflows, balancing performance, cost, and security. Responsibilities include building scalable agent architectures, automating deployments with GitHub Actions, and implementing Langfuse for observability and evaluation.

Qualificações

• Hands-on AI engineer with production deployment experience.
• Proficient in Claude Agent SDK, AWS Bedrock AgentCore, Langfuse, GitHub Actions.
• Experience building observable, scalable enterprise AI solutions.

Responsabilidades

• Design, develop, and deploy AI agents in Claude SDK and Python.
• Deploy agents on Amazon Bedrock AgentCore Runtime and integrate with data sources.
• Configure Langfuse for tracing, monitoring, and evaluation of agent traffic.
• Build CI/CD pipelines with GitHub Actions for automated deployments.
• Create AWS Step Functions workflows for document processing with Textract and Comprehend.
• Observe metrics, latency, and costs; troubleshoot production issues.

Conhecimentos

Claude Agent SDK

AWS Bedrock

AgentCore Runtime

Langfuse

GitHub Actions

Observability

LangChain

RAG

CI/CD

Python

Ferramentas

Terraform

Docker

OpenSearch

Pinecone

FAISS

pgvector

Amazon Textract

Amazon Comprehend

LangGraph

LangChain

Descrição da oferta de emprego

ZealHire Inc. is seeking an experienced AI Engineer to design, deploy, and monitor production-grade agents using Claude Agent SDK and AWS Bedrock AgentCore.

You will translate complex business needs into reliable, observable AI workflows, balancing performance, cost, and security. Responsibilities include building scalable agent architectures, automating deployments with GitHub Actions, and implementing Langfuse for observability and evaluation.