Artificial Intelligence & Machine Learning , Big Data Security Analytics , Next-Generation Technologies & Secure Development
Using Generative AI Tools to More Effectively Clean Up DataNorwest's Rama Sekhar on How Large Language Models Can Solve Data-Wrangling Issues
Artificial intelligence can solve really old problems around data wrangling and data protection that are essential to many security investigations, said Norwest Venture Partners' Rama Sekhar.
See Also: Demystifying Managed Detection and Response Services
Security data today often arrives either in logs or in an unstructured form, forcing a SOC engineer to manually go through the data and determine where the adversaries are working and what they're doing, according to Sekhar. He said Norwest is looking at emerging companies that use large language models to automatically clean up data without any manual intervention needed as possible investment targets.
"With these large language models, we can enable SOC analysts who might be a year on a job or six months on a job to do very sophisticated analysis that would otherwise require somebody who has experience for 10 or 20 years," Sekhar said. "It's changing the way we're searching, creating scripts and writing. The SOC analyst has to do a lot of this."
In this video interview with Information Security Media Group at RSA Conference 2023, Sekhar discusses:
- How SOC analysts can capitalize on generative AI large language models;
- How SIEMs and data lakes gain from AI's ability to prepare and normalize data;
- How threat actors use tools such as ChatGPT to create and execute attacks.
Sekhar focuses on early- to late-stage venture investments in enterprise and infrastructure including cloud, AI/ML, DevOps, cybersecurity and networking. His current investments include Amberflo, ClearDATA, DataRobot, Dremio, FOSSA, Harness and InfluxData.