BAST.AI stands for Artificial Intelligence in Talent Development. CEO Beth Rudden founded Bast to bring ethical AI to deployment areas with highly sensitive personal data. In 2019, we deployed her AI for the first time to create data-based insights: transparent, explainable and accessible to their owners and authorized users in accordance with German data protection law. 

Full transparency on data sources enables the owners of the data to validate it directly, creates trust and invites to share more high quality data for a transparent give-to-get. Insights are explained by referring back to the data used, right down to the word or sentence. Users are guided in natural language for full focus on results (instead of menu navigation).  

Step 1:

We take your data and extract the entities and relationships that describe your organization's language. We generate a formal knowledge graph called an ontology. 

Step 2: 

We use the ontology to understand your data in relation to the language of your organization. Listen to Beth on the importance of language.

Step 3:

We can use the ontology to infer that you may have evidence for skills or competencies.

An Ontology represents a network of information with logical relations. The picture above shows a knowledge graph of an ontological network of skills related to one specific skill that we searched (blue circle top left). All other circles are representing skills related to the target skill. Colour code, position and relation to each other is based on explainable relations. Ontologies come with 4 main characteristics: 

1. Concept search: Concepts directly related to a particular skill are used, as well as any known language variation (beyond key word searching)
2. Continously learning: Ontology development is an iterative process. It is continously improving and validating (machine and human)
3. Explainable: Ontologies function like a "brain". They work and reason with concepts and relationships in ways that are close to the way humans perceive interlinked concepts. 
4. Scalable: Easy to extend and add to existing ontologies. As a result, the model evolves with the augmenting data structure without impacting related systems.