Conversational artificial intelligence
SANIA uses language models to understand naturally phrased requests and generate conversational answers.
- Natural-language questions
- Answers built from context
- Different engines possible by configuration

SANIA combines a visible interface in the physical space with conversational artificial intelligence and the organization’s own knowledge.
For visitors, the experience should remain simple: ask a question and receive a relevant answer. Behind that interaction, SANIA connects the interface, voice or touch input, the conversational engine and the business information supplied by the organization.
The architecture can evolve according to the project. The goal is not to impose a single model, but to select components that fit the actual environment.
SANIA uses language models to understand naturally phrased requests and generate conversational answers.
Relevance also depends on the business information made available. SANIA can use a knowledge base populated with different documents and content supplied by the organization.
Interaction can be voice-based and, depending on the configuration, touch-based. The interface should be chosen according to noise, hardware, traffic and audience needs.
SANIA can be deployed on different compatible screen and kiosk formats. Hardware selection depends on location, readability, acoustics and installation constraints.
SANIA integrates with the Sandisplay digital signage ecosystem. This makes it possible to design projects where broadcast information and conversational interaction are considered coherently.
When a project needs to exchange data with a third-party system, that function must be assessed separately. A technical possibility should not be confused with a native SANIA feature.
Multiple LLM engines can be used depending on the project configuration. This makes it possible to adapt the solution to identified objectives and constraints without presenting one model as the answer to every need.
ALLSAN has worked in digital signage and interactive solutions since 2007. This real-world screen experience complements SANIA’s AI layer: placement, readability, interaction, maintenance and content consistency remain essential to deployment.
A deliberately simplified view of a question’s journey without exposing proprietary implementation details.
The visitor asks a question.
The interface receives the interaction.
The request is interpreted with session context.
The conversational engine determines how to build the answer.
The knowledge base can provide relevant business information.
A configured function, API or webhook can be called if needed.
The answer is delivered back through the SANIA experience.
Technical capabilities are enabled according to the actual requirements of the project.
SANIA can retrieve relevant information from documents and knowledge-base content.
The LLM can select a configured tool when an intent needs an action or external information.
HTTP endpoints can be called to exchange data with an external system included in the integration.
An application can expose actions that SANIA may trigger within the boundaries defined for the project.
The ability to call an API, webhook or function does not mean that every CRM, ERP, booking, payment or business system is already natively connected.
Each third-party connection must be defined, secured, tested and integrated for the relevant project. This keeps platform capabilities clear without promising an integration that has not been built.
Existing screen, dedicated kiosk, network constraints, business content or integration needs: a demonstration helps identify the most relevant SANIA configuration.