Details.
Build and own the control plane that governs MING Labs' fleet of AI agents. Responsibilities include registering, scoping, permissioning and versioning skills and tools, controlling costs and model routing, creating audit trails, enforcing decision-authority tiers and approvals, building infrastructure that takes agents from role brief to production, maintaining context and retrieval systems, enforcing least privilege and credential isolation, and owning deployment reliability, authentication, secret rotation and security gates. The role also includes pressure-testing off-the-shelf runtimes and leading the design and development of MING Labs' own agent harness. MING Labs is looking for strong platform or backend engineering experience, a governance and security mindset, comfort designing control-plane systems, evidence-first habits, the autonomy to ship unsupervised, and fluency with context and retrieval substrates. Nice to have experience includes LLM evaluation and observability tooling, agent frameworks or runtimes, and container and fleet operations at scale. The role offers real autonomy, ownership of governance and audit capabilities, a path to platform lead, high-leverage work, and flexible, location-open working.