srswti-axis (AI eXplainable Inference & Search) is an industry-ready machine learning library for advanced data retrieval and interpretable AI solutions. it provides a collection of powerful components and state-of-the-art capabilities that run entirely locally on your machine. we focused on efficient implementations that deliver high performance even on modest hardware.
For complete documentation, visit docs.srswti.com. Visit srswti.com for our products, and check out our blogblog to learn more about our mission and the mathematical foundations behind our work.
- Local intelligence—too cheap to meter, firewall-protected for privacy, on-premise steerable, and rendering benchmarks obsolete
- Everything what LLMs do but much faster, averaging 50-100ms latency
- Industry-ready
- Zero-shot classification <70ms latency
- Advanced text cleaning and preprocessing
- Intelligent text analysis and NER
- Document clustering and ranking
- Neural-probabilistic hybrid search
- Multi-strategy document merging
- Multilingual translation support
- Sentiment analysis
- Python 3.10 - 3.11
Create a new virtual environment using uv:
uv venv --python 3.10Activate the virtual environment:
For Unix/macOS:
source .venv/bin/activateFor Windows:
.venv\Scripts\activateUsing uv:
uv pip install srswti-axisOr using traditional pip:
pip install srswti-axisInside activated virtual environment:
srswti-axis-post-installthis command takes a little time at first depending on your hardware.
Note: We will be updating the notebooks section in the coming days to provide plug-and-play examples for using SRSWTI Axis. Check back soon for comprehensive Jupyter notebooks demonstrating the full capabilities of the framework.
We are optimizing our frameworks every day. Please email us at team@srswti.com or raise an issue here, and we will typically respond in less than an hour!