Founding Rust Engineer

About the role

PLOMID persists rows, documents and time-ordered events through one page-oriented storage contract covering layout, durability and recovery. It is the foundation everything above it stands on, and this role owns it. You will work in Rust on the code where a wrong byte is a lost write, which means the work rewards care more than speed.

What you’ll do

• Design and implement page, block and segment layout: what lives where on disk, and why.
• Build the write-ahead path: append, flush, commit, sync, plus the checkpoints that bound replay.
• Implement crash recovery that an operator can trust without reading the source first.
• Add CRC verification at the boundaries that matter, and make corruption loud instead of silent.
• Work with the query layer on what the storage interface promises: snapshots, scans, visibility.

What we’re looking for

• Production Rust experience, including unsafe code you can justify line by line.
• A working understanding of how databases persist data: WAL, checkpoints, page formats, buffer management.
• Comfort reasoning about durability: fsync semantics, torn writes, what “committed” actually means.
• Small, reviewable patches and design notes written before large changes.

Nice to have

• Contributions to storage engines, filesystems or embedded databases.
• Experience with MVCC, B-trees or log-structured storage.
• Familiarity with the PostgreSQL wire protocol.

What you’ll work on

The storage and wal crates: layout, the B-tree, the buffer pool, segmented WAL, checkpoints and recovery. The architecture page describes the layers this code sits under; the storage documentation describes the surface it exposes.

Why this role matters

Every query PLOMID will ever run touches this layer. A page format chosen well now pays off for years; one chosen poorly taxes every model built on top of it. This is the role where that choice gets made.