Role Overview
The Kernel team builds Neo4j’s storage engine and transaction machinery. You’ll work on low‑level systems—custom storage engines, page cache, transaction handling (WAL, ACID), indexing, and memory management—determining query performance and data safety for billions of nodes. The team is based in Malmö and London, reads database research, and runs “Do What You Want” weeks for experimentation.
Responsibilities
- Design, profile, and implement core database features, including storage formats, index structures, and recovery paths.
- Diagnose concurrency bugs, race conditions, and lock contention.
- Perform performance engineering: profile CPU, memory allocation, off‑heap memory, and disk I/O to improve latency.
- Review designs and pull requests with emphasis on correctness, data integrity, and long‑term maintainability.
- Help decide technical direction and architectural trade‑offs as the engine evolves.
Requirements
- Systems programming experience: comfortable working close to hardware. Most code is Java; C++, Rust, or Go background with memory layout and concurrency understanding is acceptable.
- Strong grasp of concurrency and data structures: threads, locks, atomic operations, cache locality, and algorithmic complexity.
- Pragmatic engineering: write thorough tests to avoid painful edge‑case debugging in production.
- Collaborative attitude: comfortable with technical choices being questioned, back up decisions with data/benchmarks, and treat code review as a two‑way discussion.