So you worked with another financial services organization on data compliance, which is a very different kind of challenge. 

What happened there? Absolutely. So in that particular project, we needed to create a way to anonymize. It was about 1.3 petabytes of data, about 13.5 billion objects. 

So that was across structured and unstructured data. So we had to create a solution that brought the data in, and then we created a LLM-based solution that identified what data needed to be anonymized, and we achieved 97.5% accuracy and 100% compliance.