Senior Backend Engineer (Python), AI Engineering: Agent Foundations

An overview of this role

As a Senior Backend Engineer on the AI Engineering: Agent Foundations group, you’ll build and evolve the core Python service that powers GitLab’s Duo Agent Platform. You’ll design and run the platform that hosts and coordinates AI agents and flows, enabling both GitLab teams and customers to automate development tasks and improve productivity across their workflows. You’ll lead efforts to apply modern Python web development practices, orchestrate communication between services via gRPC, and contribute directly to agentic AI capabilities and interactions with the latest large language models. In a company where Ruby is the primary language, you’ll be the go-to Python expert. You’ll set standards, mentor others, and shape technical and product decisions that define how AI is integrated into GitLab’s DevSecOps platform.

Some examples of our projects:

• Duo Workflow Architecture
• Componentizing Duo Workflow Graph

What you’ll do

• Lead the design and development of a highly scalable Python web service that hosts and coordinates AI agents using frameworks such as FastAPI or Starlette and gRPC.
• Develop and evolve integrations with the latest large language models, including prompt design and agentic behavior orchestration within the Duo Agent Platform.
• Build and maintain robust testing infrastructure to validate agent behaviors and ensure reliability, observability, and scalability across distributed systems.
• Drive adoption of modern Python backend best practices, setting standards for code quality, performance optimization, and security across the service.
• Provide technical mentorship and guidance to engineers across the AI Engineering and Code Creation teams, raising the overall technical bar.
• Influence product and architecture decisions by proposing iterative solutions to complex technical challenges in multi-service environments.
• Collaborate with cross-functional teams in product, engineering, marketing, and more to shape AI-powered development capabilities and customer-facing experiences.
• Participate in Tier 2 and Tier 3 on-call rotations to troubleshoot production, security, and urgent engineering issues in a globally distributed environment.

What you’ll bring

• Deep experience building and operating Python web services in a multi service environment using frameworks such as FastAPI or Starlette.
• Proven track record designing and leading complex, distributed backend systems with a focus on reliability, observability, scalability, and performance for high volume APIs.
• Hands on experience integrating AI and machine learning capabilities into products, including work with large language models and prompt design.
• Ability to architect and evolve gRPC based service interactions and collaborate effectively with teams working in other languages and stacks.
• Experience mentoring other engineers, providing technical guidance, and influencing engineering strategy across multiple teams.
• Comfortable representing engineering decisions and AI powered capabilities in discussions with product partners, customers, and other stakeholders.
• Strong written and verbal communication skills for working asynchronously in a remote, globally distributed team.
• Openness to learning new technologies and applying transferable skills, including experience in languages such as Go, Typescript, Ruby, or IDE extension development as a plus.

About the team

The Agent Foundations group sits within GitLab’s AI Engineering organization and is responsible for building the Duo Agent Platform, which provides the core services, agents, and flows that power AI-driven development experiences across GitLab. We are distributed across multiple regions and time zones, collaborating asynchronously through GitLab issues, merge requests, and written proposals to design, implement, and operate a highly scalable Python and gRPC-based platform. We partner closely with other GitLab AI teams and product groups to ensure that agentic capabilities can become the backbone for many current and future AI features, while tackling challenges such as reliable orchestration of agents, robust observability, and safe, testable AI behavior.