Senior Software Development Engineer, AWS ElastiCache

Description

This is an opportunity to join one of AWS's most foundational and high-impact engineering teams — the Data Plane team within Amazon Elasti Cache and MemoryDB. We own the core in‑memory engine that powers millions of customer workloads: the data path, durability layer, replication protocol, snapshot system, and advanced data structures. Our work sits at the intersection of systems programming, database internals, and performance engineering — every microsecond matters when you’re serving the world's most latency‑sensitive applications.

If you’ve ever found yourself deep in a conversation about write‑ahead logs, copy‑on‑write fork semantics, lock‑free data structures, or memory allocator design — and you want to apply those ideas to a system handling millions of operations per second at cloud scale — this team is where you belong. We build the engine behind Amazon MemoryDB, the only Valkey‑compatible database that delivers in‑memory speed with Multi‑AZ durability.

We’re not just running an open‑source cache; we’re extending Valkey with novel capabilities — durable replication, active‑active conflict resolution, full‑text and vector search — while maintaining the sub‑millisecond latencies our customers depend on.

Our customers include Disney+, Snap, Zoom, Lyft, Airbnb, and hundreds of thousands of other AWS customers who trust us with their most performance‑critical workloads. You’ll work in C/C++ at the lowest levels of the stack, solving problems in memory management, I/O optimization, replication protocols, and distributed data systems — all in production at massive scale.

As a Software Development Engineer on the Data Plane team, you will take on broad ownership of the core engine that sits at the heart of Elasti Cache and MemoryDB.

• Designing and building the next‑generation in‑memory database engine, delivering sub‑millisecond latencies and millions of operations per second to the world's most demanding applications.
• Developing large‑scale distributed in‑memory storage systems in C/C++, extending open‑source Valkey with durability, replication, and advanced data structure capabilities.
• Building and optimizing the durability layer — transaction logging, snapshotting, and replication protocols — that enables MemoryDB to deliver in‑memory speed with Multi‑AZ data protection.
• Designing and implementing advanced data structures and query capabilities including vector search and full‑text search to support emerging workloads like generative AI.
• Driving performance engineering at the systems level — memory allocator tuning, I/O path optimization, and lock‑free concurrency — to push throughput and latency boundaries.
• Contributing to active‑active replication and conflict resolution mechanisms that enable globally distributed, low‑latency data access.
• Mentoring and growing engineers on the team, serving as a technical leader and role model for engineering best practices.
• Managing individual project priorities, deadlines, and deliverables with a high degree of autonomy and accountability.

Day‑to‑day, you can expect a dynamic mix of deep technical work and collaborative engineering. A typical week might look like:

• Writing and reviewing production‑quality C/C++ code for the core in‑memory engine — optimizing data structures, replication paths, and the durability layer for maximum throughput at minimal latency.
• Participating in design reviews and architecture discussions, where you’ll debate trade‑offs around durability guarantees, memory efficiency, and replication consistency — and then go build the solution.
• Collaborating with peer engineers to debug complex systems issues at the engine level — analyzing crash dumps, tracing memory corruption, and profiling hot paths under production‑scale load.
• Working with upstream open‑source Valkey contributors and internal partner teams to integrate new capabilities and ensure our extensions remain compatible and performant.
• Engaging with product teams to understand emerging customer needs — from generative AI workloads requiring vector search to globally distributed applications…