RAG

Retrieval-Augmented Generation pipelines and techniques

1.

Retrieval-Augmented Generation WIP

RAG pipeline — retrieval, augmentation, and generation

2.

Document Parsing & Extraction WIP

Libraries for parsing documents and extracting structured data (Docling, LangExtract) for RAG pipelines

3.

Web Scraping & Crawling WIP

Tools for scraping, crawling, and extracting web data for AI pipelines (Spider, Playwright, Crawl4AI)

4.

Lexical Search (TF-IDF & BM25) WIP

Statistical keyword matching, term frequency, and the BM25 ranking function

5.

RAG Evaluation Metrics WIP

Core metrics for evaluating Retrieval and Generation quality in RAG pipelines

6.

RAG Evaluation Frameworks WIP

Tools for automating RAG evaluation: Ragas, DeepEval, and TruLens

7.

Advanced Retrieval & Routing WIP

Semantic routing, hybrid search (BM25 + Vector), and filtering strategies

8.

Query Expansion & Reranking WIP

HyDE, Query Rewriting, Cross-Encoders, and CRAG (Corrective RAG)

9.

GraphRAG WIP

Combining Knowledge Graphs with RAG for multi-hop reasoning and global context retrieval

10.

Vision & Late Interaction RAG WIP

ColBERT, ColPali, MUVERA, and Vision-based RAG (VisRAG) for multi-modal document retrieval

11.

Enterprise RAG Platforms WIP

End-to-end open-source RAG platforms: MaxKB, Dify, and FastGPT