Staff Backend Engineer

An overview of this role:

As a Staff Backend Engineer on the GitLab Knowledge Graph team, you'll help design, scale, and operate a high‑impact graph data service that underpins agents, analytics, and architecture‑level features across GitLab.com, Dedicated, and Self‑Managed deployments. You'll partner with a small, senior Rust‑first team to ship reliable graph capabilities and make them easy for other teams and agents to use.

The Knowledge Graph service is a distributed SDLC indexing system. It builds a property graph from GitLab SDLC (software development lifecycle) and code data using ClickHouse, NATS JetStream, and the Data Insights Platform. It also exposes secure graph queries and MCP tools for AI agents and product features. In this role, you’ll own core parts of the system end to end: shaping the architecture, hardening multi‑tenant behavior and performance, and making it straightforward for other teams and agents to consume graph capabilities.

In your first year, you’ll take clear ownership of major areas of the service (for example, the graph query engine, SDLC indexing, or multi‑tenant authorization), reduce single points of failure through better runbooks and shared context, and raise the bar on how we design, build, and operate analytical services across the stack.

What you’ll do

• Lead the design and evolution of core Knowledge Graph services in a production Rust codebase, including the graph query engine, SDLC and code indexing pipelines, and API/MCP surfaces that other GitLab teams and AI agents rely on.
• Own complex, cross-cutting initiatives that span GitLab Rails, the Data Insights Platform (Siphon, NATS, ClickHouse), and GitLab Duo Agent Platform, from technical direction and design docs through implementation, rollout, and iteration.
• Drive system design decisions that improve reliability, scalability, and maintainability for analytical (OLAP‑style) graph workloads. This includes multi‑hop traversals, aggregations, and multi‑tenant isolation. Document trade‑offs so the broader team can move quickly and stay aligned.
• Define and improve operational maturity for the service, including service level objectives (SLOs), observability, runbooks, incident response, capacity planning, and production readiness (PREP) for GitLab.com, Dedicated, and Self-Managed deployments.
• Collaborate asynchronously with product, data, infrastructure, security, and AI teams to sequence work, unblock platform‑level dependencies, and land features in a way that is safe for customers and sustainable for the team.
• Apply AI‑assisted development workflows responsibly (for example, using MCP‑aware tools, Knowledge Graph-backed agents, and internal Duo tooling) and help establish practical norms for how the team uses AI while maintaining strong engineering judgment.
• Mentor and support other engineers through pairing, technical design reviews, and knowledge-sharing, reinforcing shared ownership of the system and its operational sustainability.
• Contribute across the stack when needed, including occasional Ruby (Rails integration and authorization paths) or frontend work (for example, the Software Architecture Map UI) to close gaps and keep delivery moving.

What you’ll bring

• Significant experience building and operating production backend systems, with a track record of owning reliability, maintainability, and on-call readiness for services that support other product teams or platforms.
• Strong engineering skills in Rust or clear evidence you can ramp quickly and deliver in a Rust-first, performance-sensitive backend codebase.
• Strong system design skills, including making and explaining clear architectural decisions, documenting constraints, and aligning trade-offs with product and platform needs.
• Strong fundamentals in preparing and structuring information for AI agents: how to curate and organize what the agent sees, design systems that support effective LLM-powered behavior, and manage context windows and token usage.
• Comfort working in ambiguous environments, with the ability to work autonomously and stay self-directed. Demonstrated ability to identify problems, drive solutions, and take ownership.
• Experience with distributed data or analytics systems (for example, OLAP databases like ClickHouse or columnar stores, Kafka‑ or NATS‑style messaging, or change data capture (CDC) pipelines), and comfort reasoning about trade‑offs in that space.
• Familiarity with graph data modeling and/or query patterns (property graphs, Cypher/GQL, n-hop traversals, aggregations), or strong interest in developing that expertise in this role.
• Practical experience using AI tools in day-to-day development, with the ability to explain how you structure prompts, validate outputs, and fold AI assistance into a disciplined engineering workflow.
• A language-agnostic mindset and evidence that you can learn new languages and frameworks as the problem requires (for example, Ruby, Go, or TypeScript/Vue in adjacent parts of the stack).
• Excellent written communication and comfort collaborating asynchronously across teams and time zones in an all-remote environment.
• Interest in helping grow others through mentoring, thoughtful code review, and sharing context as the team scales and more customers adopt Knowledge Graph.

The Knowledge Graph team sits within the data engineering organization and builds the backend service that turns GitLab's SDLC and code data into a unified property graph. We expose it through a high-performance, ClickHouse-backed query engine and MCP tools. We're a small group of senior engineers working closely with partners across AI (Duo Agent Platform), analytics, infrastructure, delivery, and security because our work touches many layers of the platform. We work asynchronously and value strong ownership: each engineer is expected to build a deep understanding of the system and help evolve it over time. As we grow adoption, we're focused on scaling the service sustainably and making it reliable and easy to operate for GitLab.com, Dedicated, and Self-Managed customers.