Data Science

RAPIDSFire Podcast: Cybersecurity Data Science with Rachel Allen and Bartley Richardson

Kick off 2021 with a listen to the newest episode of RAPIDSFire: the Accelerated Data Science Podcast! 

Host Paul Mahler sits down with AI Infrastructure Manager at NVIDIA, Bartley Richardson, and Senior Cybersecurity Data Scientist at NVIDIA, Rachel Allen.

In this episode, they discuss the intersection of Data Science and Cybersecurity, specifically, how NVIDIA GPUs are helping solve problems that have plagued experts in the cybersecurity space for years. 

Log files generated by various systems in a company’s infrastructure can leave traces of attacks a data scientist can turn into actionable insight. The problem, historically, has been that logs do not conform to any standard format and the relevant information can be difficult to extract.

Traditionally, this has been done through regular expressions. Bartley and Rachel discuss how their extension of the BERT model, cyBERT, turns logs into usable data faster and more efficiently than was previously possible.

They also discuss the ready-made cybersecurity pipeline available in the RAPIDS CLX GitHub repo, super-fast GPU tokenization for NLP applications, and what the future holds for the RAPIDS CLX team. 

This podcast is a great way for anyone in the data science, NLP, or cybersecurity spaces to begin the work year inspired and engaged. Listen to the latest episode here

Discuss (0)