Ryan Reeves
Ryan is a senior software engineer at Overwatch Data, specializing in the research and development of effective human-machine teams. Before joining Overwatch Data, he worked at a cyber threat intelligence firm and contributed to offensive cyber R&D programs for the U.S. government. Interested in all things intelligence and grokking data at scale.
ryan@overwatchdata.io
linkedin linkedin.com/in/reevesrs24
mastodon: infosec.exchange/@reevesrs24
Session
Data is the lifeblood of the digital age, and its creation is proliferating at an unprecedented scale. In 2024, over 300 million terabytes of data will be generated daily, a figure that will only grow as the internet continues to permeate every aspect of our society. This data encapsulates the full spectrum of human experience, from humorous cat memes to devastating cyberattacks. Making sense of these vast streams of information, traveling at lightning speed through cables and airwaves, is a Herculean task, yet a crucial one if we are to mitigate the potential risks lurking within.
Threat intelligence analysts are inundated with data, and agentic AI systems can be invaluable tools for rapidly analyzing vast unstructured datasets, filtering out noise, and accelerating insights. But how can these agentic systems assist, and what role should
they play in the threat intelligence ecosystem? This discussion will explore the current landscape of agentic systems, the principles of their design, and the strengths and weaknesses of deploying these agents with minimal supervision in the real world.