Pathway LLM App
NormalDescription
Ready-to-run cloud templates for RAG, AI pipelines and enterprise search with live data, always in sync with Sharepoint, Google Drive, S3, Kafka and more.
Key Features
- Live data sync — automatically syncs incremental changes from file systems, Google Drive, Sharepoint, S3, Kafka, PostgreSQL and other sources
- Built-in vector index — vector search via usearch and hybrid full-text search via Tantivy, no external database required
- Multimodal RAG — GPT-4o parsing extracts charts, tables, and text from PDFs/DOCXs in real time
- Adaptive RAG — reduces token consumption by up to 4x while maintaining accuracy
- Docker containerized deployment — every template runs as a Docker container exposing an HTTP API
- Multiple templates — pre-built Q&A RAG, document indexing, multimodal RAG, unstructured-to-SQL pipelines
Use Cases
💡 Enterprise real-time search: build always-current AI search across Sharepoint, Google Drive document libraries
💡 Financial document analysis: extract structured data from PDF reports into PostgreSQL for natural language queries
💡 Multimodal knowledge retrieval: semantic understanding and precise retrieval of complex documents with charts and tables
💡 Private deployment: fully local RAG system via Mistral + Ollama for data privacy compliance
💡 Smart slide search: multimodal indexing and retrieval of PowerPoint and PDF presentations
Strengths & Limitations
✅ Strengths
- • High community interest (59.0k stars)
- • Permissive open-source license (MIT)
- • Established track record (3 years in production)
- • Responsive to issues, low backlog
Categories
Quick Start
# Install Pathway and dependencies
pip install pathway
# Run the Q&A RAG template via Docker
docker run -p 8080:8080 -e OPENAI_API_KEY=$OPENAI_API_KEY \
pathwaycom/llm-app:question-answering-rag
# Or run the Python script directly from the template directory
cd templates/question_answering_rag
pip install -r requirements.txt
python app.py