Tigris published a case study on how we built Basic Memory Cloud — per-tenant buckets, rclone sync, snapshots, and why those primitives let us preserve user ownership.
This week, Tigris posted a terrific case study about Basic Memory, our philosophies around AI memory, and why we chose Tigris as the storage layer behind Basic Memory Cloud.
As our users know, Basic Memory is built around a simple idea: your memory should belong to you. That means plain Markdown files, storage you can inspect and move, and cloud infrastructure that preserves portability instead of hiding your knowledge inside an opaque system. Tigris helps us make that work in Basic Memory Cloud by giving each user isolated, S3-compatible storage, with sync, snapshots, and file history built on top.
They’re also a smart, thoughtful company that seems to run with the same kind of energy we try to bring to our own work. When we email them, they reply. When we talk about what we wish Tigris could do, they give us great advice, and half the time we realize they already offer what we’re asking for. We just didn’t know it yet.
It was a pleasure connecting with Katie Schilling and David Myriel, who wrote the piece. They took the time to really understand Basic Memory, and their case study gets into the technical choices behind the product, including per-tenant buckets, rclone sync, bucket snapshots, and how those primitives let us build cloud features without giving up user ownership.
We’re reposting it here for readers who want a deeper look at the infrastructure choices behind Basic Memory, and the philosophy behind them: AI memory that belongs to you. If you’re at all interested in the details of AI memory, it’s worth a read.
Knowledge tools built to last
© 2026 Basic Memory - Basic Machines