On September 20, 2024, the NU Annual Research Conference 2024 at Nazarbayev University continued with the Mini Symposium on AI. The symposium began with a plenary session titled “Harnessing AI for Societal Good” by Prof. Mohan Kankanhalli, Provost’s Chair Professor, Deputy Executive Chairman of AI Singapore, and Director of the NUS AI Institute. The esteemed expert covered topics such as the AI in Health Grand Challenge, AI Prize Challenges, the Global Challenge for Safe and Secure LLMs, and the importance of fairness in AI models and generative AI. The session was insightful and thought-provoking for the NU research community and the Institute of Smart Systems and Artificial Intelligence (ISSAI) team.
The parallel session on AI Research for Societal Impact featured a panel of distinguished speakers from Nazarbayev University. The speakers included Associate Professors from the School of Medicine, Prof. Mei Yen Chan and Prof. Abduzhappar Gaipov, as well as Associate Professors from the School of Engineering and Digital Sciences, Prof. Ferhat Karaca, Prof. Siamac Fazli, and Prof. Zhanat Kappasov. Their presentations covered a wide-ranging exploration of AI’s potential benefits to society.
Additionally, the Symposium on AI included a panel session focused on generative AI Models, which attracted significant attention. This session began with presentations that laid the groundwork for detailed discussions. Madina Mansurova, Head of the Information Technology Faculty at Al-Farabi Kazakh National University, shared her experience first. Then Alibi Zhangeldin, Lead Data Scientist at NITEC JSC, presented NITEC’s ecosystem and projects. Contributions from both sessions highlighted a broad and interconnected understanding of various aspects of AI, from practical applications to the innovative field of generative models.
ISSAI team members, including Deputy Director of Product and External Affairs, Senior Data Scientist Madina Abdrakhmanova, Senior Data Scientist Askat Kuzdeuov, Soyle Product Lead Ulzhan Bissarinova, and Data Scientists Rustem Yeshpanov and Rakhat Meiramov, presented their projects.
Askat Kuzdeuov discussed “Current State of the Kaz-LLM Project: Data Preparation, Model Training, and Applications” detailing the process of data preparation and model training for the foundational Kaz-LLM model. Rustem Yeshpanov shared a presentation on “Kaz-LLM Leaderboard: A Benchmark Suite for Large Language Models in Kazakh” explaining differences between various LLMs and demonstrating benchmark examples with tricky sentence interpretations. Rakhat Meiramov talked about “Virtual Avatars”, highlighting the digital human prototype Umay, developed at ISSAI, and its potential applications in education and other fields.
Ulzhan Bissarinova introduced the Soyle product, demonstrating its text-to-speech and text-to-text translation features in Kazakh, English, Turkish, and Russian, and showcased the Soyle App to the audience. Madina Abdrakhmanova presented a work on “Rainfall-runoff Modeling via AI for Kazakhstani Watersheds”.
The ISSAI panel session attracted many attendees, including distinguished invited speakers of the NU Annual Research Conference, Prof. Andreas Cangellaris and Prof. Mohan Kankanhalli. Prof. Mohan Kankanhalli specifically shared his positive impressions of ISSAI’s contribution to AI development in Kazakhstan and thanked the organizers for the insightful and interesting symposium.