#1414th of 21 in RAG and knowledge
serviette
RAG and enterprise search app templates built on the Pathway live data framework
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 58.8k
- License
- MIT
- Last commit
- Jul 2026
- Language
- Jupyter Notebook
56,492 on 25 Mar 202658,787 on 14 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
A set of Python templates for RAG and document search that stay in sync with sources such as local files, Google Drive, SharePoint, S3, Kafka and PostgreSQL. Indexing is built in (usearch for vectors, Tantivy for full-text), so no separate vector database, cache or API framework is needed. Templates include basic Q&A, multimodal GPT-4o parsing, unstructured-to-SQL, Adaptive RAG, a private Mistral/Ollama variant and video RAG, each exposing an HTTP API.
Who it is for: Teams building live-synced RAG or document search backends in Python
Strengths
- Index updates on additions, deletions and edits in connected sources
- Built-in vector, hybrid and full-text indexes, no external vector DB
- Private template runs locally with Mistral and Ollama
- Optional Streamlit UI for quick demos
Weaknesses
- Tied to the Pathway framework, so pipeline code is not portable
- Several templates call hosted APIs (GPT-4o, TwelveLabs)
- No tagged releases
- README gives no RAM or hardware requirements
What it needs
- Needs Pathway, usearch, Tantivy, PostgreSQL (SQL template), Ollama (private template), TwelveLabs (video template), Streamlit (optional UI)
- Models: OpenAI GPT-4o, Mistral via Ollama, TwelveLabs Pegasus and Marengo
Also in RAG and knowledge
See all 21| Rank | Project | Score |
|---|---|---|
| 1 | deeptutorSelf-hosted AI tutor with knowledge bases, guided learning, books and quizzes | 78 out of 100 |
| 2 | RAGFlowRAG engine with deep document parsing, template chunking and agentic retrieval | 74 out of 100 |
| 3 | Open NotebookSelf-hosted NotebookLM alternative with podcasts and 20+ model providers | 74 out of 100 |
| #4 | LightRAGGraph-based RAG framework with a REST API server and web UI | 74 out of 100 |
| #5 | MaxKBEnterprise knowledge-base agent platform with RAG, workflows and MCP tools | 69 out of 100 |