Voice Is the New UI for Physical AI | Brendan O'Driscoll artwork

Voice Is the New UI for Physical AI | Brendan O'Driscoll

MTS

October 14, 2026

Samsara VP of Product Brendan O'Driscoll discusses the general availability launch of Agent Studio, explaining how voice agents, no-code builders, and physical AI are reshaping real-world operations in fleet management, safety coaching, and site logistics. Turn ideas into software people love.

0:00

SPEAKER_1

Hello, everyone, and welcome back to MTS. Today, we are joined live in person with Brendan O'Driscoll, visiting us all the way from his hometown in Orlando. He is the VP of Product at Samsara, and they recently, just this morning, announced the launch of Agent Studio. This brings AI agents to trucks, cabs, job sites, maintenance bays, all this cool stuff. So Brendan, welcome to MTS.

0:23

Brendan O'Driscoll

Thank you so much for having me. It's awesome.

0:24

SPEAKER_1

Yeah, thank you for joining. I think what's really interesting, your VP of Product, you've probably seen how Samsara as a platform, has evolved over its origin to what it is now. Today, Agent Studio is generally available. Tell me a little bit more about how the platform has evolved.

0:39

Brendan O'Driscoll

Yeah, it's been an exciting journey. I'm relatively new or tenured, but I think if you think about Samsara in its classic sense, it's been an enterprise SaaS products. You have all of your generic dashboards to maximize the amount of people that you can service. I think as we transition to this AI forward, this is the most generic first version of Samsara. It inverts the product. Rather than being all things for all people, just in case, you can now be exactly the right thing for exactly the right person. We're flipping the product so that you can ask it questions. We can derive what you're trying to do and turn that into automated tasks or custom dashboards specifically for you as the individual customer. That's a big change.

It reduces the surface area and makes the solutions much more tailored to the individual.

1:30

SPEAKER_1

Yeah. You can design the agent exactly for the use case that the person might actually need it. Tell me a little bit about Agent Studio. This is the thing that you guys actually officially announced today.

Tell me a little bit about what was announced.

1:42

Brendan O'Driscoll

Big day. Agent Studio came into general availability today for all of our users. What it is, it is a no-code builder for voice and chat agents. You can choose from a bunch of preset templates for common workflows. Imagine a start of shift voice agent that might brief your driver on everything they need to know or a hours of service alert agent that might keep track of all of your drivers if they're approaching their limits and where they should rest up.

I think what's really interesting outside of the templates is that you can actually just create agents using natural language as well. So imagine just talking about your operations saying, here's generally what I need to do, and every time this thing happens, I want that other thing to happen. We can just take that natural language where you've just typed or actually literally just spoken to Samsara and convert that into a series of agents that do work on your behalf.

Yeah. It's out today. You can build with it, you can test, you can set the permissions and get something live in a couple of minutes.

2:44

SPEAKER_1

It's interesting too because you had some interesting insights around how people are actually using it in the real world that may differ from how the traditional employee in an office might be interacting with chatbots. So what did you notice maybe about how the UI, like they're using voice instead of just typing all day? Like what are you noticing there?

3:02

Brendan O'Driscoll

For us, the big insight is that voice is the new UI for these agentic workflows, right? So intuitively, I think it makes sense. We're operating in the real world, imagining drivers driving trucks, you know, they can't use their hands, or an operator walking through a site from A to B, or operating some machinery.

It doesn't really make sense to kind of sit and wait behind a prompt in a chat interface. We can actually meet the customer where they are and use voice agents, which are just much more intuitive of an interface. So we have opportunity to make both chat agents and voice agents. And since we've launched both, we've seen the voice agents just massively accelerate and outpace the chat agents.

And ultimately, I think that's what's interesting about working in the real world is, is we can meet the customer where they are and the way that's easiest for them. And for us, that's voice. So voice has a big role to play going forward in this space.

3:55

SPEAKER_1

100%.

I'm hearing this consistent theme about meeting customers in the real world where they actually are. I want to get your experience because you're VP of product. You talk to these customers all the time. Give me some concrete use cases about how maybe the real world person, like the truck driver or someone at the maintenance site, is actually using these agents as opposed to maybe how the office employee might be using agents.

Read the other 13 minutes of this episode here.

$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.

Not what you expected? Full refund within 14 days.

Thousands of transcripts fetched by people building searchable podcast archives

Fetch transcripts with code

The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000796274795