template
RAG Template Pack
5 production-ready RAG pipelines. Each one is a complete, deployable implementation, not a tutorial.
What's included
- ✓PDF RAG: chunk, embed, retrieve, rerank. Works with any PDF.
- ✓Web RAG: crawl, index, query. Built on Playwright + ChromaDB.
- ✓SQL RAG: natural language to SQL. Handles joins and aggregations.
- ✓Multi-modal RAG: text + images in the same index.
- ✓Conversational RAG: multi-turn with context compression.
- ✓Evaluation suite included for each pipeline
- ✓OpenAI + Anthropic + Cohere embedding support
Code sample
python
from rag_templates import PDFPipeline
pipeline = PDFPipeline(
embedding_model="text-embedding-3-small",
llm="claude-sonnet-4-6",
chunk_size=512,
)
pipeline.ingest("docs/manual.pdf")
result = pipeline.query("What's the return policy?")