Read storage, database, and distributed-systems papers and turn them into reusable design constraints, assumptions, mechanisms, and positioning. Use when surveying a topic, extracting implementation ideas, comparing systems, or building related-work and novelty framing.
Run an end-to-end workflow for storage, database, and distributed-systems research ideas: scout hotspots, mine directions, test whether the problem is real, search prior art, judge feasibility, assess novelty, estimate likely venue level, and decide whether to invest in implementation or writing. Use when actively looking for topics, triaging a new idea, or deciding whether an existing prototype is publishable.
Review storage, database, distributed-systems, distributed-file-system, and object-storage designs or code for correctness, durability, recovery, membership safety, metadata discipline, and lifecycle robustness. Use when auditing a PR, design doc, incident, test plan, or research roadmap involving replicated state, storage metadata, namespace service, control-plane logic, or distributed data movement.
Implement or refactor single-node storage-engine and database internals such as LSM trees, B-trees, WAL, MVCC, iterators, compaction, checkpoints, snapshots, file formats, metadata indexes, and hot read/write paths. Use when writing low-level code or tests where invariants, durability, and lifecycle safety matter more than API polish.
Design and interpret trustworthy performance studies, failure injection, recovery tests, and profiling runs for storage, database, and distributed systems. Use when making latency or throughput claims, comparing baselines, validating optimizations, or proving a fix survives load and faults.
Plan and steer storage, database, and distributed-systems research from problem framing to claim scoping, invariants, experiments, and milestones. Use when starting a new topic, turning engineering pain into a research question, decomposing a large agenda, or deciding what to build and measure first.
Write storage, database, and distributed-systems papers, design notes, evaluation narratives, and rebuttals from real implementation and evidence. Use when turning a system or prototype into a workshop or conference paper while keeping claims bounded to what the artifact and experiments actually support.