Senior Staff Systems Engineer - Performance Engineer - JVM

About the role

The Systems Performance team is part of the Computing Squad, consisting of two distinct workstreams—Orchestration and System Performance—each with its own approach and challenges on managing and improving the foundational infrastructure where the majority of Nubank's workloads run. The Performance team focuses on building deep diagnostic tools and performing high-level analysis to reduce latency, infrastructure costs, and increase service efficiency.

You will be responsible for:

• Leading Deep-Dive Investigations: Conduct high-level performance analysis to identify and resolve systemic bottlenecks across our global JVM-based microservices architecture.
• Optimizing Resource Efficiency: Drive initiatives to reduce infrastructure costs and latency by fine-tuning JVM parameters, Garbage Collection (ZGC, G1), and memory management (heap and off-heap).
• Building Diagnostic Tooling: Develop and implement advanced observability tools using eBPF, JFR, and Flamegraphs to provide real-time insights into kernel and runtime behavior.
• Bridge the Gap Between the Linux Kernel and the JVM: Optimize thread scheduling (CFS/EEVDF) and manage resource isolation (cgroups/throttling) within our Kubernetes environment.
• Architect Scalable Solutions: Design and deliver innovative infrastructure improvements that address long-term performance challenges, ensuring our systems scale ahead of demand.
• Technical Mentorship & Culture: Share expertise on JVM internals and performance best practices with the wider Engineering team, fostering a culture of technical excellence and "quality over quantity."
• Root Cause Excellence: Deep dive into complex concurrency issues, lock contentions, and memory leaks, providing definitive fixes for high-impact technical debt.
• Strategic Collaboration: Work closely with the Computing Squad to align orchestration strategies with system performance goals, ensuring a seamless interface between infrastructure and workloads.

Qualifications

• Expertise in JVM Internals: Deep, low-level knowledge of the JVM is essential. Must understand how the JVM works "under the hood," including JIT compilation (C1/C2), class loading, and intrinsic methods.
• JVM Tuning & Garbage Collection: Extensive experience with GC algorithms (ZGC, G1, Shenandoah), including the ability to tune them for massive heaps and ultra-low latency requirements.
• OpenJDK Contribution (Major Plus): Previous experience contributing to the OpenJDK project or other low-level runtime environments is a significant advantage.
• Linux Kernel & Scheduling: Deep understanding of the Linux Scheduler (CFS/EEVDF), thread scheduling, and how the kernel manages high-concurrency Java workloads.
• Mastery of Heap and Off-Heap Memory Management: Mastery of heap and off-heap memory management, including Direct Buffers, memory-mapped files, and diagnosing complex memory leaks.
• Advanced Diagnostics: Mastery of diagnostic tools such as Flamegraphs, JFR (Java Flight Recorder), eBPF, and performing large-scale heap dump analysis.
• Resource Isolation: Extensive experience with cgroups and the impact of CPU Throttling on JVM quotas within Kubernetes/EKS.
• Concurrency: Proven ability to diagnose and resolve complex concurrency problems, including lock contention and race conditions at the instruction level.
• Cloud Platforms: Knowledge of AWS infrastructure and its performance characteristics.

Benefits

• Opportunity of earning equity at Nu
• Medical Insurance
• Dental and Vision Insurance
• Life Insurance and AD&D
• Extended Maternity and Paternity Leaves
• Nucleo - Our learning platform of courses
• NuLanguage - Our language learning program
• NuCare - Our mental health and wellness assistance program
• Extended Maternity and Paternity Leaves
• 401K
• Saving Plans - Health Savings Account and Flexible Spending Account
• Work-from-home Allowance
• Relocation Assistance Package, if applicable.

Location

Palo Alto, United States

Work Model

Hybrid 2-3 times/week: Our hybrid work model brings us to the office at least twice a week, on strategic days designed to maximize team connection and collaboration.