ragleap-rag
io.github.antonyrag
io.github.antonyrag:ragleap-rag is a Maven Central artifact (jar) for the JVM. It has 9 published versions. Browse versions below for dependency snippets and security details.
Add to your project0.13.0
<dependency>
<groupId>io.github.antonyrag</groupId>
<artifactId>ragleap-rag</artifactId>
<version>0.13.0</version>
</dependency>Version Details — 0.13.0
Java port of ragleap-rag's RAG pipeline: connection pooling, embeddings (7 providers), vector storage on six backends (pgvector, FAISS, Qdrant, Weaviate, Milvus, Pinecone; Milvus and Pinecone are not live-verified; sparse and hybrid RRF search on pgvector), retrieval, generation (streaming, structured output, provider fallback, vision captioning), conversation memory, query rewriting, and ingestion: a document parser for 27 of the Python package's 28 formats (no Parquet; legacy .xls needs the optional Apache POI dependency), OCR through the Tesseract binary, audio extraction from video through the ffmpeg binary, audio transcription (OpenAI Whisper, Deepgram or a custom function; not live-verified), and an ingestion service that ties them together - plus standalone utilities (chunking, sanitization, guardrails, cost tracking, evaluation scoring, structured-output validation, observability hooks). Independently versioned from the Python ragleap-rag package (see https://pypi.org/project/ragleap-rag/) rather than kept in lockstep. The RagLeap entry point (ask, streaming, retrieve, evaluate, history, document methods, batch ingestion) is included; URL ingestion, reranking, the Redis query cache and the async variants are not ported. Every module here is genuinely tested; documented per-class where it is not live-verified.