HeadlinesBriefing favicon HeadlinesBriefing.com

Antfly: Go-Based Distributed Search Engine with Built-in ML

Hacker News •
×

Antfly is a distributed document database and search engine written in Go that combines full-text, vector, and graph search capabilities. The single-binary deployment includes native ML inference through Termite, eliminating the need for external API calls for vector search. Developers can use it for multimodal search across text, images, audio, and video.

Built on etcd's Raft library with CockroachDB's Pebble storage engine, Antfly features a multi-Raft setup where metadata and data shards get separate consensus groups. The system supports MongoDB-style in-place updates, streaming RAG, and automatic relationship extraction for knowledge graphs. A Kubernetes operator and MCP server enable enterprise deployment and LLM tool integration.

The project ships with Elastic License v2, allowing modification, self-hosting, and product building while prohibiting managed service offerings. This licensing choice balances sustainability with open-source principles. Antfly includes pgaf for PostgreSQL integration, React components for search UIs, and Termite for built-in ML operations like embeddings, chunking, and reranking.