Artifact is a semantic query platform based on generative artificial intelligence, which allows users to access and consult structured knowledge bases, such as knowledge graphs, simply by asking questions in natural language.

 

Traditionally, querying a knowledge graph requires the use of SPARQL, a technical language and protocol that is often unfamiliar and therefore inaccessible to many developers. Artifact eliminates this barrier: it interprets users’ natural language requests, automatically translates them into SPARQL queries, executes them and returns clear and readable answers, both in structured (e.g., tabular) and discursive form.

 

This is achieved through the adoption of a neuro-symbolic approach, where traditional generative AI is hybridised with the Semantic Web, ensuring accurate and reliable responses and thus minimising the hallucinations typical of large language models. Thanks to this combination, Artifact allows you to interact with data in a natural and intuitive way in chatbots, voice assistants or other conversational channels, facilitating the exploration of relationships between data in different application domains (e.g., culture, public services, environment, healthcare).