Head of Platform Engineering

You are accountable for two integrated outcomes: effective engineering productivity — velocity that survives contact with production, not velocity for its own sake — and production reliability — the SRE posture, rollback guarantees, and incident response that let teams move fast without breaking payments. These are not two jobs. They are one job, because the platform is the mechanism through which both are delivered.

Own the IDP. Standardized service templates, CI/CD pipelines, deployment tooling, feature-flag infrastructure, progressive delivery, automated rollback. Golden paths for each tech stack we currently operate, faster to use than to work around. A platform-as-a-product discipline where dLocal engineers are your customers.

Own SRE and production excellence. Reliability engineering across dLocal's payment platforms. Observability standards. Incident management maturity. Production readiness reviews. On-call health. Error budgets. The operational contract between platform and product engineering teams.

Own AI-assisted engineering, done right. You own how AI shows up in the developer workflow at dLocal. That means the evolution of dCoder — our proprietary AI code generation platform built on advanced models from Anthropic, OpenAI, and others — the integration of third-party assistants where they add value, the prompts and context engineering that keep AI output consistent with our architectural standards, and the guardrails that stop AI-generated code from reaching production without the same review, testing, and observability we require of any other change. You will ship tools, teach teams how to use them, measure the outcomes, and iterate.

Own standards and architectural consistency. You partner with our Principal engineering community and the Engineering Bar Raiser Program to define and enforce the standards that let dLocal scale without fragmenting. Code review, testing, deployment, and AI usage standards.

Drive organizational influence and adoption. Owning the platform is 30% of the job. The other 70% is changing how hundreds of engineers work. You will lead the adoption program: enablement, internal comms, champions, migration plans, deprecation discipline. You measure success by how many teams are on the paved road, not by how many roads you have paved.

Deliver against the metrics that actually matter. Not vanity numbers. Deployment frequency and lead time for changes reported alongside change failure rate, time to restore, and code turnover rate — the four traditional DORA metrics plus the AI-era additions. SLO attainment on critical payment paths, rollback success rate, paved-road adoption, AI-assisted PR defect density versus human-written baseline. A deployment frequency win that comes with a CFR regression is not a win.