Synthetic data typically focuses on structured datasets, but this session explores how organizations can generate realistic synthetic text to complement that data while preserving privacy. By combining structured data with curated narrative examples and leveraging large language models, SAS enables the creation of context-rich, varied text such as emails or incident reports. This approach improves model training, supports rare event simulation, and enhances testing as systems move from development to production. Integrated with SAS tools like Visual Text Analytics and agent workflows, it provides a scalable and governed way to generate privacy-safe narrative data.
Watch the recording
