VAST Data

So before we reveal our sources, here's this episode's three-point briefing.

Number one, today's guest, Jeff Denworth, has spent more than two decades in the data world, with leadership stops at DDN and Cetera before co-founding Vast Data in 2016.

Number two, today, Vast Data is valued at $30 billion.

That's the most valuable private data infrastructure software company in history.

And number three, we're coming to you from the W Hotel in San Francisco, which, strangely enough, has kept active beehives on its rooftop since 2011.

I'm your host, Kevin Cochrane, Chief Marketing Officer here at Vulture.

And we are broadcasting from San Francisco in the AI Conference, where I happen to have the pleasure of sitting down in the...

W Hotel having a literal fireside chat with a nice hot fire behind me with Jeff Denworth, the co-founder of Vast Data.

They can search data, they can make decisions, they can take action.

But for enterprises, this raises a basic problem: how do you let agents work without giving them too much access? And how can enterprises make sure agents aren't losing track of what they're doing? To talk about this and so much more, today I'll be sitting down with Vast Data Co-founder and CTO Alon Horev.

From agent sandboxes, security, observability, and frankly, the infrastructure needed to put agents to work safely at scale, we will be covering quite a bit of ground, so stay tuned.

[upbeat music] So Alon, we've gone from an era where everyone's using their own individual, uh, AI tools, so you're using Claude or you're using ChatGPT to perform certain functions or tasks.

Uh, if an agent needs to query an analytical system that gets updated, you know, once an hour or once a, once a day, it may lead to, you know, um, less optimal outcomes.

Uh, so being able to, to address that challenge of how can we enable real-time access to any type of information, be it unstructured, like a video stream or an audio stream, uh, but also for structured information, which is sensor readings or, um, whatever, you know, analytical information that's coming into a system, um, I think that's a massive opportunity that we definitely see customers are interested in, but then it requires reimagining how to store, uh, that kind of information.

Um, and we're doing our best to make it very easy to leverage the vast database in order to store that type of information and stream it in real time, uh, into agents.

Um, I think another interesting aspect of where agents are going, um, is around physical AI.