Multi-agent dbt orchestration at Riot Games: redefining the analytics engineering SDLC

At Riot Games, we are moving beyond AI as a "text generator" to AI as an "Agentic Colleague." And we're using dbt to ensure we can support safe, multi-agent orchestration to redefine our data SDLC in the AI era.

Using dbt agent skills (now a part of dbt Wizard) and the dbt Fusion engine, we built a development workflow where agents securely read metadata for automated discovery, translate complex legacy logic into dbt models, and generate PRs for code review. The whole thing runs within a read-only guardrail architecture that keeps humans in control at every step.

In this session, we'll walk through our architecture in detail: how we scoped agent permissions to read-only, what the guardrails look like in practice, and where human review stays non-negotiable. You'll leave with a concrete blueprint for deploying multi-agent dbt workflows safely, and a clear picture of what "safe" actually requires in a production data SDLC.