AI/ML Engineer — MCP Infrastructure

Role overview Own the infrastructure that keeps a production agentic AI system reliable as its servers, tools, and workloads grow. You’ll work with a small engineering group, taking responsibility for deployments, delivery pipelines, observability, and operational stability.

Responsibilities - Monitor and improve infrastructure running agent frameworks and related services. - Manage containerized deployments and keep environments stable. - Build and maintain CI/CD pipelines for agent and model releases. - Set up tracing, alerts, and performance dashboards for LLM-based systems. - Help integrate MCP servers as new components are developed. - Find and address latency, usage cost, and reliability problems before they cause incidents.

Requirements - Practical DevOps experience owning deployments, using Docker and a container orchestration service such as ECS or Fargate. - Experience building CI/CD workflows with GitHub Actions and working with AWS services such as ECS, RDS, S3, and CloudWatch. - Familiarity with the MCP protocol and production experience with LangChain, LangSmith, or LangGraph. - Understanding of production LLM infrastructure, including token costs, latency tradeoffs, rate limits, and monitoring. - Ability to work independently in a small team and communicate fluently in English with US-based colleagues.

Nice to have - Experience configuring LangSmith traces and evaluations and interpreting their results. - Familiarity with observability tools such as Datadog or Grafana. - Experience working directly with a senior engineer on a client team.

Benefits and work setup - Remote contract work from Latin America, aligned to US Eastern hours, Monday through Friday, 9 a.m.–6 p.m. - Monthly compensation of $4,000–$5,000 USD, paid biweekly. - Initial six-month contract with potential to extend; paid time off accrues after a three-month trial period.