About the Role
This role focuses on building an AI-first software delivery system where autonomous agents handle end-to-end feature development, while engineers design and maintain the safe, scalable platform that enables it.
Responsibilities
- Design and implement autonomous workflows that transform requirements from specification through to code, review, and deployment
- Create self-improving agent loops that convert errors into systemic enhancements rather than isolated corrections
- Develop evaluation frameworks for both offline testing and continuous integration to catch behavioral regressions beyond code correctness
- Establish and manage review checkpoints combining automated systems and human oversight for high-risk changes
- Enhance the dependability of tools through typed interfaces, defined schemas, retry logic, timeouts, and safety validations
- Build infrastructure for persistent agents, including session management, state boundaries, memory controls, and recovery mechanisms
- Advance the platform's architecture using TypeScript and serverless patterns to prioritize simplicity and long-term maintainability
- Collaborate with Product teams to clarify ambiguous requirements and convert them into verifiable, testable specifications
Work Arrangement
Remote — Australia
Team
Small team of 3–4 engineers reporting to a hands-on CTO, with company-wide adoption of agentic workflows across functions.
Why this role exists (and why it’s different)
Six months ago, 'agentic engineering' seemed experimental. Now, it's quickly becoming standard. While most companies use AI to speed up coding, we're creating an AI-first delivery model where agents independently ship features. Engineers in this role build the platform—the 'harness'—that ensures safety, reliability, and scalability. If successful, product growth scales with ambition, not team size.
What we mean by “harness engineering”
An agent is more than a language model—it's the model plus the surrounding system. In this role, you'll build that system: agent execution environments and loops (plan → act → observe → reflect, with retries and stop conditions), feedback systems that turn results into improvements (evaluations, regression tracking, learning), tool integrations (MCP, APIs, secure credentials, sandboxes), governance layers (permissions, policies, human-in-the-loop checks, audit logs), observability tools (traces, cost tracking, failure classification, response guides), and evaluation systems (scenario testing, trajectory scoring, tool argument validation, non-deterministic unit tests).
What success looks like in ~12 months
A production-ready managed agent platform trusted by product teams for customer-facing features, agent workflows that deliver consistently instead of requiring manual fixes, well-defined quality controls including evaluation systems, automated reviewers, regression testing, and rollout safeguards, and engineers spending less time on direct reviews and more on improving system reliability, speed, and safety.
Who this role is not for
Be honest: if you prefer spending most of your time building end-user features directly, this role will feel unsatisfying. If you're enthusiastic about AI but lack experience shipping real software, you won't have the judgment needed to assess agent output. If you thrive in environments with clear scope, established practices, and low ambiguity, this fast-moving, experimental setting won't suit you.
The team and company
You'll join a small engineering team of 3–4 people reporting to a hands-on CTO. The entire company—not just engineering—is adopting agentic workflows across sales, marketing, and support. This isn't a side project; it's the core operating model.
Additional Information
Our culture is built on trust, respect, and ownership, guided by values: Embrace Change, Carte Blanche, Find Wisdom in Data, and We All 'Own It'. Fully remote within Australia. 5 weeks of annual leave with flexible hours. Monthly wellness allowance for mental and physical health. All team members receive employee share options (ESOP).
How to apply?
Submit your CV through the application form. Background checks are part of the offer process. We welcome applicants from all backgrounds, abilities, and identities. Diversity strengthens our creativity, innovation, and success. Help us build a workplace where everyone can thrive.
Other
- Background checks are required as part of the offer process
- Fully remote (work from anywhere in Australia)
- 5 weeks annual leave and flexible working
- Monthly Wellness Budget (mental & physical health)
- Employee share options (ESOP) for all team members
- We welcome applications from all backgrounds, abilities, and identities
- We value diversity and believe that it enhances our creativity, innovation, and overall success
