Senior Data Engineer — LLM Agent Orchestration

OpsBrasil Serviços Cloud LTDA; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

OpsBrasil Serviços Cloud LTDA seeks an experienced software engineer to drive production-grade LLM agent projects. You will own design, tooling, and deployment from requirements through client discussions, focusing on orchestration, tool calls, and secure write actions.

You’ll operate across AWS Bedrock, LangGraph, and Terraform with strong emphasis on cost, latency, and observability. The role requires hands-on Python, LLM design skills, and the ability to defend technical trade-offs during

Qualifications

• 5+ years in software engineering, with at least 2 at a senior level.
• Hands-on LLM application engineering: prompt design, tool use, structured output.
• Experience with agent orchestration beyond single-prompt calls.
• Production IAM, networking, storage, and observability in cloud environments.
• Ability to defend design decisions directly on client calls.

Responsibilities

• Own end-to-end delivery: from requirements to production deployment and client defense.
• Build and harden agent orchestration and routing, tool-calling, planning and state management.
• Integrate tools (BM25, vector/hybrid search) for scalable retrieval.
• Run models on AWS Bedrock and Bedrock AgentCore with guardrails and memory management.
• Create evaluation harnesses and instrument with OpenTelemetry for per-session metrics.
• Optimize cost and latency with routing, caching, pruning, and circuit breakers.
• Design safe write-actions with least-privilege permissions and feature-flag releases.

Skills

Python production-grade

LLM application engineering

Prompt design

Orchestration & tooling

System design & architecture

Tools

AWS Bedrock

LangChain

LangGraph

Bedrock AgentCore

Terraform

OpenTelemetry

CI/CD (GitHub Actions)

Job description

OpsBrasil Serviços Cloud LTDA seeks an experienced software engineer to drive production-grade LLM agent projects. You will own design, tooling, and deployment from requirements through client discussions, focusing on orchestration, tool calls, and secure write actions.

You’ll operate across AWS Bedrock, LangGraph, and Terraform with strong emphasis on cost, latency, and observability. The role requires hands-on Python, LLM design skills, and the ability to defend technical trade-offs during