We use cookies (and other similar technologies) to improve your experience on our site. By using this website you agree to our Cookie Policy

We built a novel graph topology, to overcome the scalability limitations of traditional graphs. We use graph as a fundamental building block - a ground-truth for language processing models.

As humanity, we are far away from AGI, but at Smabbler we believe that the broad adoption of graphs with their abilities for connecting objects, carrying features and populating multidimensional relations, will help us build solutions that are more context-aware.
The way Galaxia maps data at a basic level conceptually refers to ontologies, a representations of knowledge with concepts (nodes) and relationships between them (edges).
We are confident that our work will eventually lead to the wider availability of a global ontology used to help improve every knowledge-based AI solution.
Unlike solutions based on neural networks, which leverage large amounts of data to train AI systems to perform requested tasks, our graph works with unique data points.
We don't need to crawl the internet, we don't need users' data.
Text models
Our text models utilize an advanced way of inducing connections between graph nodes. The initial application of Smabbler Galaxia is text recognition and extraction with a high level of unification and accuracy.
When testing the capabilities of our graph, we confirmed its application in creating a powerful, yet very light, embeddings system, and we are determined to soon start working on the productization of Galaxia-powered embeddings.

Smabbler Galaxia ™

Technology overview whitepaper

Thank you! Request granted.
Oops! Something went wrong .