Agentic Developers / Architects

HMG AMERICA LLC; United States; Remote; USD 170,000 - 210,000; Full time

Job summary

HMG AMERICA LLC seeks Agentic Developers / Architects to advance Agentic AI systems from POC to production on AWS AgentCore. Candidates should demonstrate multi-agent orchestration, containerized deployment, and scalable reliability across environments.

The role emphasizes MCP-compliant tooling, secure interpreters, and robust integration with AWS services for production workloads. Remote US East Coast preferred along with New York tri-state area proximity.

Qualifications

• Proven track record taking AI agents from POC to production.
• Experience designing and implementing multi-agent orchestration workflows.
• Familiarity with security and resilience in containerized environments.

Responsibilities

• Architect and implement agent orchestration pipelines for production use.
• Design standardized APIs using MCP and ensure secure execution runtimes.
• Collaborate with platform teams to deploy and operate AgentCore-based solutions.

Skills

Agentic orchestration

AgentCore/AWS

MCP standards

Tools

LangGraph

Strands SDK

AWS AgentCore

Model Context Protocol

Job description

Location: New york- tri state area - Remote (US East Coast preferred)

Agentic Developers / Architects

A critical prerequisite for these roles is a proven track record of taking Agentic AI systems completely from POC to production environments, specifically leveraging AWS Agent Core and related architectures. No Domain requirement.

Key technical competencies we are looking for include:

• Production Orchestration (LangGraph / Strands SDK): Experience building complex, cyclic, multi-agent workflows where agents hand off tasks to one another, handle exceptions, and recover from failures without losing state.
• AWS Bedrock AgentCore Platform Engineering: Hands-on experience deploying containerized agent execution loops via AWS AgentCore, configuring the AgentCore Gateway for tool discovery, and managing production-grade compute isolation.
• Enterprise Tooling & Protocol Standards (MCP): Mastery in designing standardized tool APIs using the Model Context Protocol (MCP), alongside securing automated execution runtimes like sandboxed Python Code Interpreters.