The role
Supabase is building a software-defined NVMe-oF storage platform beneath its Postgres fleet, and you'd own it from architecture to production. The first goal is raw I/O performance, with snapshots and branching to follow.
What you'll do
- Lead architecture of the NVMe-oF storage platform, including build-vs-buy choices
- Integrate networked block storage into Kubernetes via CSI drivers and controllers
- Profile and tune the full Postgres I/O path, from fsync and WAL to the network
- Design durability, failover, I/O isolation and QoS across tenants
- Set SLOs, mentor engineers and join on-call rotations
What they're looking for
Must-haves
- 7+ years of software engineering, including production storage systems
- Built or operated distributed block storage such as Ceph RBD or Portworx
- Hands-on NVMe/TCP or NVMe-oF, including targets, initiators and multipath
- Deep Linux storage internals: block layer, io_uring, O_DIRECT, device-mapper
- Strong systems code in Go, plus C, C++ or Rust
Nice to have
- SPDK development experience
- RDMA/RoCE experience
- Postgres internals knowledge or familiarity with OrioleDB
The company

Supabase
Remote-first open-source Postgres platform
Supabase is a born-remote, open-source-first Postgres development platform with about 400 team members across 60+ countries. It has raised over $1B, including a $500M Series F.
Summarised by Jobs That Travel from Supabase's listing.







