Principal Software Engineer (Roblox Database)
Posted Updated
- Roblox’s database team develops the next-generation, multi-tenant database platform that elastically scales and underpins every online data workload at Roblox
- As a principal engineer on the database team, you will shape the architecture, build and launch critical database capabilities that keep our services fast, reliable and efficient at global scale
- You will report to the Technical Director for Storage
- Design and implement new engine features—indexing, storage formats, WAL and replication protocols, sharding, and query-planner enhancements—that push latency, throughput, and availability boundaries
- Evolve the control plane to deliver elastic scaling, autonomous healing, and zero-downtime schema or tenant moves across global regions
- Profile and optimize critical code paths using kernel-level tracing and advanced performance tooling; drive systematic tail-latency reductions
- Establish engineering best practices by leading design reviews, performance benchmarks, failure drills, and post-incident retrospectives
- Automate everything: develop frameworks for testing, CI/CD, rollout safety, observability, and autoscaling so that the platform operates hands-off at scale
- Mentor and unblock engineers, fostering deep database craftsmanship and disseminating knowledge across Storage, Platform, and product teams
- Engage with the open-source community—upstreaming patches, presenting learnings, and influencing the future of Postgres, RocksDB and related ecosystems
- Passion for automation, rigorous testing, and data-driven decision-making that raises the reliability and efficiency bar
- 8+ years of experience in software engineering or a relevant field
- Deep expertise in database internals—B-trees, MVCC, WAL, query optimization, buffer management—and strong skills in C/C++, GO or Rust
- Ability to turn ambiguous product requirements into a clear technical roadmap and influence cross-functional stakeholders
- Bonus: upstream contributions to PostgreSQL, RocksDB or other OSS databases; familiarity with disaggregated storage, HTAP workloads, or custom query languages
- Hands-on experience with consensus algorithms (Raft, Paxos) and multi-region replication models that deliver strong consistency and high availability
- Proven success shipping high-QPS, low-latency services on container orchestration stacks (Kubernetes or Nomad) with comprehensive observability
- Deep experience with building and operating large-scale distributed databases or storage engines (e.g., PostgreSQL, MySQL/InnoDB, RocksDB, Spanner, Aurora, AlloyDB)