Imagine having a super-smart digital assistant that listens to a meeting and instantly translates different languages so everyone understands each other. It doesn't just transcribe words; it can summarize the whole conversation and answer your questions about what was decided. It's like giving every employee a personal secretary who is fluent in every language and never forgets a detail.
Hybrid and global meetings suffer from language barriers and a lack of efficient documentation. Current AI tools are often too English-centric, prone to hallucinations, and lack the emotional or cultural awareness needed for high-quality customer support.
The project built multimodal XR models for transcription, translation, and summarization. It specifically created two prototypes: a personal meeting assistant and an advanced global customer service assistant.
If you are a global consulting firm dealing with hybrid meetings across different time zones and languages — this project developed XR models that provide transcription, translation, and automated minuting. This allows teams to collaborate in their own languages while maintaining a clear, shared record of decisions.
If you are an international online store dealing with a diverse global customer base — this project developed an advanced customer service assistant. It helps human agents provide support with better clarity, empathy, and cultural awareness in the customer's own language.
If you are a legal agency dealing with complex, long-form multilingual discussions — this project developed confidence-aware and explainable models. This ensures that summaries and transcriptions are robust and the user knows why the AI made specific decisions.
Based on available project data, no specific pricing or cost structures are mentioned; the project focuses on developing the technology and tools for European take-up.
Yes, the project specifically aims to make these technologies scalable and develop efficient methods to deploy large models in an energy-efficient manner for production.
Based on available project data, the project uses a cascaded grant programme and releases tools to facilitate the use of pre-trained XR models across Europe.
The project targets the 'next level of support' for online and hybrid meetings, building upon the capabilities of existing video conferencing tools like audio, screensharing, and chat.
The project runs from 2022-10-01 to 2025-09-30, with annual releases and evaluations of use-case prototypes.
The consortium is well-balanced for technology transfer, consisting of 5 partners across 4 countries. With an industry ratio of 40% (including 2 industrial partners and 1 SME), there is a strong bridge between the academic research led by Universiteit van Amsterdam and commercial application.
Contact Universiteit van Amsterdam regarding the UTTER project deliverables.
Contact us to explore how to integrate these XR transcription models into your customer service workflow.
