All Tools
B
DataFreeOpen Source
BADGER
Fast, embeddable key-value DB for Go
Apache-2.0
ABOUT
Go-based AI services and data pipelines need fast local persistence without running a separate database server or embedding a C/C++ library. Existing options require CGO or impose Java runtime overhead. Badger solves this as a pure Go embedded key-value store that delivers high throughput with LSM-tree-based storage, ACID transactions, and configurable compression. It runs in-process with no external dependencies, supports concurrent reads and writes, and handles datasets that exceed available RAM through efficient disk-backed storage — making it ideal for Go-based ML services, feature stores, and data pipeline components.
INTEGRATION GUIDE
1. Store and retrieve feature embeddings and model metadata locally in Go-based inference servers
2. Power persistent storage for stateful stream processing in Go-based data pipelines
3. Cache intermediate results and pre-computed values in Go ML training and serving infrastructure
4. Serve as the embedded database for Go applications needing ACID transactions and crash-safe storage
5. Store configuration state, lookup tables, and model registries in minimal-dependency Go microservices
TAGS
key-value-storeembedded-databasegostoragelsm-treedata-infrastructure