VectifyAI/PageIndex

PageIndex = RAG without vectors Pipeline shift: Docs → Tree Index → LLM reasoning → Answer Not: Docs → Chunks → Embeddings → Similarity search For long docs (finance/legal), it’s hitting ~98% accuracy #Rag #pageindex

Matched through a repo-specific project phrase query.

Built Pebble: a manual-to-checklist AI workflow. - local LLM inference with Docker Model Runner (Gemma 4 4B). - vectorless RAG with PageIndex. Stack: React, Express, FastAPI, Redis/BullMQ, MongoDB Github Link: github.com/SAYOUNCDR/Pebble #LLM #Gemma #Docker #RAG #BuildInPublic

Matched through a repo-specific project phrase query.

Thanks for sharing PageIndex and our bet on "reasoning-as-retrieval"! github.com/VectifyAI/PageInd…

Returned by a high-confidence repo query and contains a visible project phrase, but the exact URL, slug, or package name was not visible.

🚀 Just executed: PageIndex: Vectorless, Reasoning-based RAG. Document Q&A — no vector DB, no embeddings, no chunking. 📊 98.7% accuracy on FinanceBench vs ~85-90% for traditional RAG. 🔗 github.com/rkfshakti/PAGEIND… #RAG #LLM #AI #OpenSource #GenAI

Matched through a repo-specific project phrase query.