XARA – XR Dialog Trainer

XARA is a customised dialogue simulator based on artificial intelligence and integrated with OpenAI (GPT-4). The application enables realistic and dynamic dialogues. Scenarios and characters can be customised without restriction via the XARA web platform. Thanks to the use of AI, every conversation takes a different course, with the virtual characters responding individually to the user's questions. This enables an interactive and realistic simulation of conversations. The application is accessible both as a VR and AR application and via a web UI.

Features

• AI-based speech processing (GPT-4)
  
• User-defined dialogues
  
• Unique, individualised conversation history
• Valuable training experiences and spontaneous reactions 
• Choice of different avatars
• Available on three platforms (AR/VR/Web)
• Downloadable conversation protocols

Scenarios and characters

Scenarios and characters/roles can be freely defined in the Xara web platform. Individual guidelines, CVs, minutes of old conversations can serve as the basis for developing a scenario. Examples of current scenarios: 

• Recruitment: Practical simulations for effective application management.
• Conflict dialogue: Targeted training for dealing with conflict situations.
• Notice of dismissal: Sensitive practice of employee dismissals for professional communication.
• Open questions: Promoting critical thinking and spontaneous reactions through open dialogues.

Benefits: Adaptability, diversity and efficiency

• Adaptability to the course of the conversation promotes flexible communication skills.
• Personalised scenarios enable targeted training for specific professional requirements.
• The uniqueness of each conversation strengthens self-confidence and the ability to react spontaneously.
• The variety of roles creates a stimulating and motivating learning environment.
• Protocol documentation enables efficient learning and improves job performance.
• Endless possibilities for new roles maximise the investment in training and increase the benefit to the organisation.

Next Steps

• RAG system: Use data from websites, documents, etc. for personalisation
• Use speech recognition/multilingual voices
• Select custom models (GPT-3.5, GPT 4, Mistral)