https://github.com/swagfin/ollamatest
RAG Demo App with Local Embeddings and Qdrant
https://github.com/swagfin/ollamatest
ai artificial-intelligence machine-learning ml qdrant qdrant-vector-database text-embeddings
Last synced: about 1 month ago
JSON representation
RAG Demo App with Local Embeddings and Qdrant
- Host: GitHub
- URL: https://github.com/swagfin/ollamatest
- Owner: swagfin
- License: mit
- Created: 2025-08-03T23:29:27.000Z (about 1 year ago)
- Default Branch: master
- Last Pushed: 2025-09-18T19:43:13.000Z (11 months ago)
- Last Synced: 2025-10-28T06:46:08.128Z (9 months ago)
- Topics: ai, artificial-intelligence, machine-learning, ml, qdrant, qdrant-vector-database, text-embeddings
- Language: C#
- Homepage:
- Size: 19.5 KB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
# RAG Demo with Local Embeddings and Qdrant
This project demonstrates how to build a simple Retrieval-Augmented Generation (RAG) system using:
- **Local Text Embeddings** via [SmartComponents.LocalEmbeddings](https://github.com/dotnet/smartcomponents)
- **Vector Storage & Search** via [Qdrant](https://qdrant.tech/)
- **Ollama** to serve a local language model (e.g., `mistral:7b`)
## 💡 Why this project?
The biggest challenge in RAG systems is **text embedding** — converting meaningful text into high-dimensional vectors for retrieval. While many solutions rely on cloud APIs or heavy frameworks like ONNX, this project demonstrates how to:
- Use **local embeddings** with minimal setup
- Achieve **fast and accurate semantic search**
- Keep everything **offline and privacy-preserving**
## ⚙️ Technologies Used
| Component | Tool/Library |
|-------------------|----------------------------------------------|
| Embeddings | [SmartComponents.LocalEmbeddings](https://www.nuget.org/packages/SmartComponents.LocalEmbeddings/) |
| Vector DB | [Qdrant](https://qdrant.tech) (running locally) |
| LLM | [`Ollama`](https://ollama.com/) (`mistral:7b`) |
| API Layer | ASP.NET Core Web API |
## 🧠 How It Works
```
[User Question]
↓
Local Embedder (Microsoft's smartcomponents)
↓
Qdrant Search (Vector DB)
↓
Top Document Retrieved
↓
Mistral via Ollama (LLM)
↓
Final Answer
```
## Setup
1. Clone the repo.
2. Install Qdrant locally or run via Docker.
3. Run Ollama with:
```bash
ollama run mistral
```
4. Launch the ASP.NET API project.
## License
MIT License.