Machine learning engineer in local

Описание:
At PPRO, we simplify access to local payments - providing a local payments platform that empowers giants like Pay Pal, Stripe, and Microsoft to connect with billions of consumers worldwide. We are on an ambitious, fast-paced mission to 3x our growth by 2028 - and we’re just getting started.

• Drive the multi-quarter roadmap for ML-driven authorization optimization, routing intelligence, and retry strategies.
• Identify the highest-impact opportunities before they become obvious, and build the case for investment with data and business framing.
• Design and lead the implementation of shared ML infrastructure—feature stores, model serving platforms, experimentation frameworks—that accelerates every team building on payments data, not just your own.
• Author and champion ML engineering standards across PPRO (model governance, monitoring, MLOps patterns), ensuring consistency, reliability, and reproducibility organization-wide.
• Take on challenges where the problem itself isn't well-defined. You scope, structure, and sequence the work—then lead execution across multiple engineers and teams to deliver.
• Actively invest in the growth of Senior engineers: through design reviews, pairing on hard problems, sponsoring stretch opportunities, and raising expectations for what "production-ready ML" means at PPRO.
• Design the experimentation strategy for live payment traffic—including multi-armed bandits, causal inference approaches, and traffic-splitting frameworks—ensuring sound statistical methodology across the team.
• Run design reviews, set expectations for technical documentation, and create the internal forums (guilds, working groups, RFCs) that help ML practitioners across teams learn from each other and align on standards.

• Demonstrated experience designing and shipping ML systems that serve multiple products or teams—not just models, but the platforms, contracts, and abstractions that make ML reusable and reliable at scale.
• Proven ability to drive technical decisions across teams you don't manage—through clear writing, credibility, and the ability to synthesize competing perspectives into a coherent path forward.
• Expert-level command of classical ML (XGBoost, LightGBM, calibration, cost-sensitive learning) with the judgment to know when—and when not—to reach for more complex approaches.
• You've operated beyond standard accuracy metrics and can design evaluation frameworks appropriate to the problem.
• Extensive experience taking models from experimentation to high-throughput, low-latency production environments.
• You've owned reliability, SLAs, and incident response for ML systems, and you've built MLOps tooling—not just consumed it.
• You write and review code at a senior+ level in Python, hold the team to high standards for testability and maintainability, and can credibly engage in systems design discussions with Principal and Staff engineers across Data and Platform.
• You connect technical decisions to business outcomes—approval rate improvements to revenue, latency reductions to conversion, model drift to operational risk.
• You communicate clearly with non-technical stakeholders and can translate ambiguous business goals into concrete ML problems.
• Strong understanding of the card payment lifecycle, issuer behavior, authorization codes, retry logic, network rules, and 3DS.
• You use domain knowledge to inform feature design, model architecture, and experimentation strategy—not just as background context.
• Deep experience designing and owning ML infrastructure on AWS or GCP at scale, including infrastructure-as-code, cost management, and the ability to make build-vs-buy decisions on platform components.

• We offer a hybrid structure with a 3 days / week on site expectation, so you can strike the…