Over the first two years of ELOQUENCE, partners across Europe have been working to turn the vision of trustworthy, multilingual conversational AI into concrete, real-world solutions. In this Partner Perspectives interview, the University of Novi Sad reflects on its key achievements so far, from building a carefully designed healthcare dialogue corpus to advancing speech technologies in Serbian. Their work highlights how responsible AI can support critical public services, while balancing innovation with reliability, transparency, and societal impact.
Q: Looking back at the first two years of ELOQUENCE, which achievement from your team are you most proud of?
We have successfully created a data corpus in cooperation with expert medical staff, consisting of recordings of telephone conversations between our team members and medical nurses working in local healthcare institutions. The dialogues simulate real interactions that may occur between healthcare professionals and callers seeking medical advice regarding their young children, including newborns. We are proud of this achievement, given the challenges involved in defining a methodology for creating a corpus that is both realistic and representative in terms of the range of topics covered, while still being controlled enough to avoid excessive spontaneity that would complicate speech recognition.
Q: What aspect of the project’s final year are you most excited about and why?
We are excited to see our work implemented in a pilot framework, where a chatbot collects case information from the caller before concluding the dialogue. In the full system, a medical professional is expected to take over at that point and provide advice, using the chatbot’s input as a strong starting point. The pilot incorporates a range of advanced NLP technologies, including speech recognition and synthesis in Serbian, the pilot language. This is particularly exciting for us, as most team members have extensive experience developing speech technology systems in Serbian.
Q: Which project result or development do you believe has the greatest potential for long-term impact beyond ELOQUENCE?
In our experience, call centres like the one simulated in the pilot are extremely valuable to society but are often limited by a shortage of medical staff, leading to long wait times and user frustration. The idea of supporting such centres with AI that collects information from callers and presents it in a clear, concise summary for the medical professional taking over the call could make much more efficient use of the expert’s time and effort. This is why we believe that call centres of this type, enhanced with ELOQUENCE technology will provide parents and child carers with timely, reliable information exactly when they need it most. We expect that such centres will become widely used and remain an essential resource for many years to come.
Q: From your perspective, what do you see as the most important opportunities and responsible AI challenges emerging right now? How do you think ELOQUENCE can help address them?
AI presents an opportunity to reduce administrative burdens and enable more efficient work and communication – but realizing these benefits responsibly requires addressing challenges such as bias, misinformation, transparency, and privacy. By focusing precisely on these challenges, ELOQUENCE emphasizes balancing innovation with accountability, a defining challenge of our time.
Q: If you had to describe ELOQUENCE’s journey so far in one sentence, what would it be?
Developing cutting-edge technology, ELOQUENCE balances innovation with accountability.
Looking ahead to the final year of ELOQUENCE, the University of Novi Sad’s contributions demonstrate how research-driven AI can meaningfully support healthcare systems under real-world constraints. By combining advanced speech and language technologies with a strong focus on responsibility and human oversight, their work points toward a future where AI enhances, rather than replaces, expert decision-making. As ELOQUENCE moves from research to pilots, these efforts underscore the project’s broader ambition: developing AI systems that are not only technically advanced, but also trustworthy, inclusive, and genuinely useful to society.
