Senior AI Systems Engineer — Agentic Software & Infrastructure

About Hyatus Living

Hyatus Living is a technology-enabled hospitality company building the operating infrastructure for modern furnished housing. We combine software, automation, and hands-on operations to deliver reliable stays at scale.

The Role

We are hiring a Senior AI Systems Engineer to build the systems through which AI agents plan, implement, test, review, and maintain production software.

This is not a prompt-engineering or autocomplete role. You will design agent loops, coding-agent factories, context and tool infrastructure, evaluation systems, and issue-to-PR workflows that produce reliable engineering outcomes.

You will work across Codex, Claude Code, Cursor, MCP, GitHub, CI/CD, cloud services, and dedicated infrastructure. A central part of the role is deciding what should run interactively, in cloud-agent environments, or on persistent Linux machines—and building the controls that make those systems dependable.

What You’ll Own

• Build agent loops that move from issue to implementation, validation, pull request, review, and merge

• Create parallel coding-agent workflows using isolated branches, worktrees, and execution environments

• Design context, memory, tool permissions, retries, handoffs, and human-review gates

• Develop automated tests, evals, observability, and failure-recovery loops

• Establish safe PR automation and auto-merge policies

• Compare and integrate Codex, Claude Code, Cursor, Warp, MCP tools, and emerging agent platforms

• Decide when workloads belong locally, in managed cloud-agent services, or on dedicated infrastructure such as Hetzner

• Build production systems using TypeScript, Python, APIs, GitHub, Linux, and cloud infrastructure

• Help engineering and operations teams achieve substantially more leverage from AI

What We’re Looking For

• Strong software-engineering fundamentals and experience shipping production systems

• Hands-on experience building with coding agents beyond individual chat sessions

• Experience designing multi-step or long-running agent workflows

• Fluency with Git, pull requests, CI/CD, testing, Linux, and cloud infrastructure

• Strong judgment around autonomy, permissions, verification, security, and human oversight

• Ability to diagnose why an agent workflow fails and improve the system—not merely rewrite the prompt

• Clear communication and comfort operating with significant ownership

Especially Relevant Experience

• Coding-agent orchestration or software factories

• MCP servers and tool integrations

• Automated issue-to-PR pipelines

• Agent evals, tracing, and observability

• GitHub Actions and auto-merge systems

• Sandboxed or isolated agent execution

• TypeScript, Python, Next.js, AWS, Docker, and Hetzner

• Codex, Claude Code, Cursor Cloud, Warp, or comparable agent platforms

How to Apply

Tell us about the most advanced coding-agent loop or software factory you have built. Include what triggered the workflow, how context and tools were provided, where it ran, how retries and validation worked, and what failed before you made it reliable.

Generic descriptions of using AI to write code will not be enough.