Ledger

An ORM that emits each database’s own dialect instead of a lowest common denominator: DISTINCT ON and JSONB operators on Postgres, ON DUPLICATE KEY UPDATE on MySQL, BSON aggregation pipelines on Mongo, PREWHERE and SAMPLE on ClickHouse. On the insert benchmark it runs about 9% over raw database/sql, against 111% for Bun and 156% for GORM.

One write path for applications that outgrew a single database. A command commits state, a versioned event and an outbox row in the same Postgres transaction; a leader-elected relay publishes to Redis Streams; the graph (FalkorDB) and search (Elasticsearch) projections are rebuilt from that log rather than written to directly. Tenancy is structural, using row level security, a graph per tenant and index routing, rather than a WHERE clause you can forget.

A declarative query language for Go. You describe what you want as a document; DQL plans it, pushes down what the database can do and finishes the rest in memory. Partition scoping is declared rather than guessed at, because getting it wrong is a data leak.

One object-storage API across local disk, memory, S3, GCS, Azure and SFTP, with middleware you stack on the read and write paths for encryption, compression or virus scanning. Content-addressed storage with reference counting and GC, and an io/fs.FS view so a flat bucket reads like a directory tree.