How do you train an AI model when the necessary training data is missing? This talk dives into the journey of developing a multi-task voice classification model and the challenges of data annotation.
Dr. Jonas Freiknecht will showcase how the creation of a mobile data annotation platform enabled community-driven open data collection, bridging the gap between AI model development and real-world data constraints.

Keeping up with constantly evolving energy regulations and assessing their impact on business processes is a major challenge. Manually analyzing complex legal texts is time-consuming and prone to errors.
This talk presents a case study with Open Grid Europe (OGE), exploring how Large Language Models (LLMs) can automate regulatory text analysis, significantly reducing manual effort.
The focus is on a human-in-the-loop approach, empowering experts to efficiently review and validate the LLM's findings, ensuring accuracy and control while leveraging the power of AI. This approach enables OGE to quickly adapt to regulatory changes and maintain compliance.