This repository showcases real-world examples of how to build Retrieval-Augmented Generation (RAG) pipelines using Moorcheh — a powerful semantic vector store designed for high-performance document retrieval and contextual AI applications.
Each folder contains a full example pipeline, complete with:
- Preloaded datasets & queries
- Code to upload and index documents into Moorcheh
- Semantic queries and retrieval logic
- Output analysis and results
| Folder | Description |
|---|---|
finance |
Semantic search over quarterly earnings reports and transcripts from major tech firms |
healthcare |
Semantic search over clinical and scientific documents related to diseases |
legal |
Semantic search over legal documents and reports |
geography |
Semantic search over geospatial data and documents |
scientific_journals |
Semantic search over research publications |
upcoming |
Other examples such as education, customer-service are in progress |
You can watch all our example demos here:
🔗 Moorcheh YouTube Channel
git clone https://github.com/moorcheh-ai/moorcheh_examples.git
cd moorcheh_examplesSet your Moorcheh API key:
export MOORCHEH_API_KEY="your_api_key_here"Before running the demo, make sure to upload your own document set (PDFs, text files, or markdown) into the designated folder—usually ./documents/—and your queries in CSV format (with a query column) into ./queries/.
cd AnalyzingFinancialDocuments_WithLangChain
python demo.pyMost examples are designed to work in both local environments and Google Colab. Within each folder, you will also find videos about each topic to guide you through the process as well.
Check out our benchmarking github and the discussion post on how Moorcheh compares with other vector databases across metrics like relevance, context completeness, and response depth.