System Development Engineer

System Development Engineer (Contractor) DRAM Memory Optimization - JDK Migration & Heap Right-Sizing Onsite: Seattle, WA Level: L5-L6 Equivalent Job Summary: As a member of the Region Flexibility Migration (RFM) team, you will be responsible for driving Amazon's DRAM memory optimization initiative focused on reducing the memory footprint of Java-based services through JDK version migrations and heap right-sizing. You will plan, execute, and validate JDK upgrades (e.g., JDK 17 to 21, JDK 21 to 25) across service fleets, resolve migration blockers, and implement optimal heap configurations that leverage new JVM features to achieve meaningful DRAM savings. This is a high-impact, independently driven role where you will navigate the complexities of JDK version migrations, including deprecated API handling, module system changes, garbage collector evolution, and bootstrap/build system compatibility, while simultaneously right-sizing heap allocations to eliminate memory waste. You will collaborate with service teams and RFM engineering leadership to validate changes, ensure zero-downtime migrations, and establish reusable frameworks for fleet-wide JDK optimization. Key Responsibilities: Plan and execute JDK version migrations (JDK 17 to 21, JDK 21 to 25) across service fleets: identifying and resolving blockers including deprecated/removed APIs, module system incompatibilities, reflection access restrictions, security provider changes, and build system (bootstrap/toolchain) compatibility issues. Perform heap right-sizing analysis for Java services: profiling heap utilization patterns, analyzing GC logs, identifying over-provisioned heap allocations, and implementing optimal Xms/Xmx/metaspace configurations that minimize DRAM consumption while preserving service performance SLAs. Evaluate and implement new GC algorithms and JVM features introduced in target JDK versions (e.g., ZGC generational mode, Shenandoah improvements, compact object headers) to achieve memory efficiency gains beyond simple heap tuning. Perform custom operations and iterative experiments using Amazon internal tooling to validate optimization impact: own end-to-end deployment, test execution, metric validation, and derive actionable insights from results. Troubleshoot JDK migration issues including class loading changes, serialization incompatibilities, JNI/native library compatibility, and performance regressions introduced by JVM behavioral differences across versions. Collaborate with service teams to review service architectures, discuss findings, propose migration plans, and align resolution strategies while communicating effectively across engineering leadership and technical stakeholders. Monitor service health metrics and troubleshoot operational issues during and after migration and heap optimization activities, ensuring zero degradation to service latency, throughput, and availability. Develop comprehensive operational runbooks, SOPs, documentation, and technical specifications that capture JDK migration patterns and heap optimization strategies, consumable by both human engineers and AI agents to orchestrate workflows at scale. Basic Qualifications: Bachelor's degree in computer science, Engineering, or equivalent technical field. 5-7+ years of hands‑on experience with Java/JVM-based services in large-scale production environments, including at least 2 major JDK version migrations (e.g., JDK 8 to 11, 11 to 17, 17 to 21) with demonstrable expertise in resolving migration blockers and compatibility issues. Deep expertise in JVM memory management, including heap sizing (Xms, Xmx, metaspace), garbage collection algorithms (G1GC, ZGC, Shenandoah), GC tuning, native memory tracking (NMT), and the ability to analyze heap dumps and GC logs to identify optimization opportunities. Strong knowledge of JDK version differences and migration challenges, including module system (JPMS) enforcement, deprecated/removed APIs, strong encapsulation of internal APIs, security provider changes, and build toolchain (Maven/Gradle/Brazil) compatibility requirements. Proficiency in Java (5+ years' experience) with strong understanding of JVM internals, including class loading, JIT compilation impact on memory, object layout, and how framework choices (Spring, Coral, Guice) impact heap utilization. Proficiency in using generative AI tools and assistants as part of daily engineering workflows to accelerate problem-solving, code development, and technical analysis. Preferred Qualifications: Direct experience with heap right‑sizing or JVM memory optimization in production distributed systems, with measurable DRAM savings achieved through GC algorithm selection, heap tuning, or off‑heap migration strategies. Experience with Amazon internal tools including Amazon Profiler, CloudWatch, X‑Ray, and load/stress testing frameworks for validating JDK migration impact and heap optimization under production‑like conditions. Familiarity with JDK 21 and JDK 25 features relevant to memory optimization, including virtual threads (Project Loom) memory implications, compact object headers, generational ZGC, and string deduplication improvements. Experience developing AI agents or automated workflows that can orchestrate tasks, extract information from services, and coordinate optimization activities across multiple systems. Strong attention to detail and effective communication abilities: able to present findings, propose strategies, and influence service team stakeholders and engineering leadership. Familiarity with CI/CD pipelines, deployment automation, and Amazon deployment technologies and best practices for infrastructure changes and migrations. #J-18808-Ljbffr ManpowerGroup Global, Inc.