October 7, 2025—Artificial intelligence (AI) systems rely on data to learn how to perform effectively during deployment. Poor-quality data can cause biased model behavior, unexpected failures, and poor model performance. The Software Engineering Institute (SEI) is bringing together experts in AI and data science to present perspectives on AI Engineering best practices for dataset generation, exploration, preparation, and testing for ensuring data quality when training AI systems. The workshop AI Data Quality: Advancing AI Engineering for Reliable Data Pipelines will be presented as a free ZoomGov webinar on October 23 from 12:30-4:30 p.m. EDT. Registration to participate is now open.

Presentations will focus on

• impacts of data quality issues on AI system behaviors
• best practices in data-centric AI
• procedures for testing data quality
• ensuring alignment between training data and the deployed environment
• identifying and removing biases in data

Paroma Varma, Snorkel AI’s cofounder and head of solutions, will keynote the workshop. Snorkel AI specializes in programmatic data development for AI.

Register to attend the AI Data Quality Workshop. Learn more about the SEI’s AI Engineering work.