Achieving robust AI model performance in industry applications is a significant challenge due to the complexity and diversity of real-world scenarios. Data-centric AI development can be used to provide the depth and nuance required for high-performing models, addressing challenges such as edge cases and bias with diverse human data. This session will cover best practices for developing a robust data pipeline for real-world data collection and annotation for building high-performing automotive AI systems. Attendees will gain practical insights into the development of cost-effective, high-quality data, including methods to address privacy and regulatory requirements.

In this session, you will:

• Discover strategies for building robust data pipelines with diverse real-world data
• Understand how diverse high-quality data can be used to develop more robust and reliable AI systems
• Learn how real-world data collection and annotation enhance datasets for DMS while ensuring compliance to regulatory requirements
• Explore cutting-edge applications for customization of LLMs in the automotive sector