Senior Backend Engineer, Trusted Agentic Development

About the Role

Join the Trusted Agentic Development team to build the future of AI-assisted software development, focusing on ensuring the reliability and trustworthiness of AI-generated code.

Responsibilities

• Own the technical thesis for validating AI-generated code and translate it into a robust architecture.
• Prototype and build feedback mechanisms to bridge the gap between passing tests and functional software.
• Develop language-aware validation approaches for Rust, Ruby, and Go, ensuring a shared definition of 'trusted'.
• Design and implement systems for mid-loop feedback to enable AI agents to self-correct.
• Collaborate with engineering teams to drive adoption, measure impact, and continuously improve the platform.
• Provide technical direction, review designs, and mentor team members.

Requirements

• Deep, hands-on experience building software with AI agents, understanding the validation gap.
• A strong perspective on closing the validation gap with concrete, tested ideas.
• Experience designing distributed systems, preferably for Kubernetes.
• Ability to design flexible systems that can be general or language-specific.
• Proven ability to prototype, learn, and ship solutions.
• Skill in influencing engineers across an organization through technical merit and clear communication.
• Excellent written communication for an all-remote, asynchronous environment.
• Alignment with GitLab's values: collaboration, iteration, and transparency.

Nice-to-Haves

• Experience with quality assurance or test automation frameworks is not the primary focus, but experience in these areas could provide valuable context.

About the Team

The Trusted Agentic Development team, part of GitLab's Developer Experience group, is dedicated to creating platforms and practices that empower engineers to confidently validate AI-generated software. We collaborate across teams and languages to understand agent usage, test new validation strategies, and establish a universally adopted model for trusted agentic development.