Website:
bluemachines.ai
Job details:
Blue Machines
About Blue Machines
Blue Machines builds the platform for production AI agents — agent builder, real-time voice agents, workflow orchestration, integrations, and inference. Multi-tenant systems, non-deterministic behaviour, many providers in the call path, and high bar for what ships to customers.
About The Role
We are hiring a founding Quality Engineering Lead — a builder who designs and ships the systems that decide what is safe to release. Your output is platforms and mechanisms, not test cases. You may grow a team later; you are hired to build first.
What You Will Do
- Design and build the test platform: execution, isolation, test-data lifecycle, reporting, flake handling, and a clear answer to whether a build should ship
- Build an agent evaluation harness that detects behavioural regressions after prompt, model, tool, or config changes
- Own release gating: enforceable gates in code, post-deploy verification, canary criteria, and a narrow hotfix path
- Treat capacity and load as release criteria, not optional reports
- Turn customer-visible defects into durable detection (owned, staged, measurable)
- Push testability into product systems — seams, replay, fault injection, observability
- Raise the engineering bar through review and mechanisms, not by writing every test yourself
What we are looking for
- Strong software engineering background — you can design a system from an ambiguous problem and implement the hard path
- Experience owning quality, reliability, release, or evaluation platforms (not only writing suites)
- Experience with conversational AI, voice, or LLM evaluation — or a clear track record evaluating non-deterministic systems without ground truth
- Release engineering, progressive delivery, or SRE depth — gates, canaries, and production signal you have owned
- Hands-on building of CI/CD quality infrastructure at meaningful scale
- Product mindset — you tie quality work to customer-visible risk and delivery outcomes
- Evidence over assertion — you quantify, and you separate known from assumed
- Ability to work with a small team and influence engineers who do not report to you
This role is not
A manual QA role, a checklist sign-off role, or a hire to write Playwright/pytest cases as the main job. We already have deep automation capability; we need the architect who builds what sits above it.
Click on Apply to know more.
