Senior Agentic Harness Engineer

We are looking for an experienced Senior Agentic Harness Engineer to work on a production-grade AI-DLC Agent Harness for an AWS-native software delivery platform.

Responsibilities:

• Build and operate agentic workflows for design, coding, review and testing.
• Implement LLM orchestration, model routing, retry and fallback mechanisms.
• Manage token usage, context windows, budgets and circuit breakers.
• Implement AI governance, safety guardrails, kill switches and audit logging.
• Integrate the agentic harness with JIRA/Atlassian APIs.
• Build monitoring and observability for agent health, cost and quality.
• Work closely with the engineering team on architecture, implementation and knowledge transfer.

The core requirements for the job include the following:

Must-Have Skills:

• Production experience building Agentic AI systems.
• LangChain / LangGraph or equivalent.
• LLM orchestration and multi-model routing.
• Model fallback and retry design.
• Human-in-the-loop workflows.
• AI safety guardrails and prompt-injection mitigation.
• Token cost and context-window management.
• Strong Python engineering.
• AWS experience; Bedrock/AgentCore is a plus.
• Langfuse, Datadog or equivalent.
• JIRA / Atlassian API experience.

Nice to Have:

• AWS Bedrock AgentCore.
• Temporal.
• LiteLLM.
• Claude Code / Claude Cowork.
• Cursor or other AI-native IDEs.
• Self-hosted open-weight LLMs.
• Financial Services or regulated-industry experience.