I wanted to share a project I’ve been working on called SemRoute, and I would love to get your feedback. SemRoute is a semantic router that uses vector embeddings to route queries based on their semantic meaning. It’s designed to be flexible and easy to use, without the need for training classifiers or using large language models.
Here’s a brief overview:
- Supports multiple embedding models (OpenAI, MistralAI)
- Support custom embedding models
- Offers dynamic and static thresholding
- Allows different scoring methods (individual averaging, centroid)
You can check it out on GitHub: https://github.com/HansalShah007/semroute. I’d really appreciate any feedback or suggestions from this community on how to improve it or use cases you think it might be suited for.
Thanks!