Applied AI Engineer

About the role

Scale's Global Public Sector team builds AI that runs inside production systems used by national health and government organisations. As an Applied AI Engineer you will own scoped components of those systems, such as a model integration, agent tool, API, data pipeline or evaluation, from implementation through to delivery. Senior engineers set the system-level design and client commitments, so you can focus on making your components work reliably on real data.

Key responsibilities

• Own the implementation, testing and delivery of defined AI components within an established system design.
• Build agent tools, model integrations and the backend services and APIs they depend on.
• Run and extend evaluations within an established framework, interpret results, investigate common failure modes and flag regressions.
• Debug production issues in the components you own and resolve them using the team's engineering practices.
• Communicate progress and blockers early and take responsibility for delivery dates within your scope.
• Work directly with client teams on requirements and feedback, and share what you learn with the team.

Required profile

• Experience building and shipping production software, with Python and working knowledge of at least one of TypeScript, Java or C++.
• Experience integrating models or AI/ML components into a production application, such as calling LLMs, building retrieval or agent tooling, or deploying a trained model.
• Understanding of ML evaluation: reading results, spotting regressions and identifying common causes of model failure.
• Experience owning a defined piece of work through testing, release and production debugging.
• Strong problem-solving skills and a data-driven approach to iterating on models and datasets.
• Clear written and verbal communication, including raising blockers and trade-offs early.

Nice to have

• A backend or full-stack engineering background.
• Experience with cloud ML platforms such as AWS, GCP or Azure.
• Exposure to agentic workflows, retrieval-augmented generation or context engineering.
• Experience handling regulated or sensitive data.