Kaizen: Using computer use and AI to give any website an API — hotel voice agents and legacy portals are first customers

Stay mo. Uh while we bring in the first uh Oh, we're ready. Let's do it. Fantastic. I thought we were going to have new news. What up? What up? Welcome to the show. I'm John. Welcome. Nice to meet you. Nice to meet you. How you doing? What's going on? Break it down for us. Break it down for us.

How's How's demo day going? Grab a water. Uh introduce yourselves. Introduce your company, please. All right. Ken, hold up the microphone a little bit. Both of you guys. Good. Ken, cool. CEO of Kaizen. Nice. Michael, CTO. Okay. You guys pitched like 5 minutes ago or 10 minutes ago? Five minutes ago.

We came right down here to spread the word. Yeah. How were the How were the nerves? Did you guys was that like a walk in the park? Yeah, it was fun. It was fun. It was You've done it at alumni and then and then probably like three to five full partnership pitches before, right? Yes. Um so I mean I've probably Yeah.

said this a h 100 times, a thousand times almost a Well, you got to say it one more time. Give us the high level. You don't have to give us the whole pitch, but break it down. All right. So Kaizen helps developers instantly integrate into websites without APIs.

You know, we work with companies across logistics, healthcare, and financial services to integrate to a wide variety of legacy portals. Very cool. Talk to me about how you're doing that. Is there MCP involved? Are you kind of using AI and LLMs to read the HTML kind of reverse engineer an API from the front end?

What's going on? 100%. The latter of what you said, uh, computer use has changed the game. A lot of this stuff, you know, you think about it.

Hey, there's a so such a small subset of software that has APIs that people can integrate with, build products on top of, but now with computer use, computers can do anything, anything available on the internet. And that's what we help companies do. Wait, reactions to 03 Pro? Have you tested it yet?

Is it uh is it improving things or do you build on open source? Like what do you like? What are you excited about in the in the AI race at the foundation model level? Yeah, I mean we love using all of them. We're friends. We're we we're Switzerland, right? Google Google. We have them all on the show.

Please let the fox in the hen house. We're Switzerland. We're Switzerland. Yeah.

No, I mean truly like different models are different for different things like you know clicking on an item on the page like we like the computer use model from Enthropic for that right pulling data from a very large table Gemini they come and clutch a little bit the biggest bigger context window exactly 100%.

So I mean that like our approach we abstract all this away from the end users they don't want to think about it they just it just happens. So, walk me through some of those use cases. Like, what's the first customer that you've had that has been like, "This solved my problem. This is amazing.

" Walk me through like a very concrete demo. One of our most f one of our favorite customers to talk about is, you know, they're a voice station for hotels. Okay.

So, in the Rio, the hotel in Vegas, you know, you call down, you say, "Hey, I want a a burger up to my room or I don't know what kind of fancy stuff you all order, but yeah, you'll talk to a voice agent. Uh, the voice agent will take your request, right? And then they'll use like calling to to get like tooth.

Last night we were staying at the Rosewood, of course. And uh we call to get toothpaste and we're like it's like it's you know you got to I don't even want to wait like you know I don't know 20 seconds while it rings and oh let me transfer you here and all that stuff.

You should just be able to pick up toothpaste please. And it just and it just comes. Exact. So how how does your service integrate with that experience? Cuz I know I pick up the phone. I'm talking to a voice agent.

I imagine you're using voice APIs to actually mediate that, but then you're designing that interaction, right? Or rather, our customer is traversing. Maybe I want to shout them out. They are a voice agent for hotels.

Uh and then they use us to write the data back into the property management system, create that service order. So they a toothpaste goes right up to your room. So instead of talking to a person and they're trying to understand you, all this stuff, it just happens instantly. What were you guys doing before this?

We were I was head of engineering at a company called Truck Smarter. I've worked with small trucking companies my whole career. Love them. Great job. Big truck guy. Big truck guy. Has lots of experience working with legacy systems, web forms. Exactly. That's that's where it comes from.

Like every shipper in the United States is a completely different website. I must have written dozens of these. I managed teams. I wrote hundreds uh and probably a lot of web scraping. But you know, now you're on and we're building Kaizen so that no one has to do that again. I love it. Talk to us about the metrics.

Did you share a headline number of users arr something to get the investors excited today? Yeah, I mean we're in the hundreds of thousands of dollars of AR started working on this. Yeah, let's blow. We got the We got all the acronyms, all this the you know what the A16Z you guys you guys got term sheets yet?

We're we're close. We have the first term sheet of the TV show. Congratulations I love this. Spreading so much debris across this room. It's terrible. Anyway, thanks for playing along with us, guys. This is awesome. We're super excited for you. That's awesome. Um, what's next on the buildout? How big is the team?

I imagine you're raising money. You're going to scale that or are you building like a much smaller team? How are you thinking about that? Yeah, I mean, this is a massive opportunity and we're excited to sprint after it. We actually have an employee right now. Yeah.

Crazy enough to to join us during the bash with the 500k in the bank account and we're we're we're doing a work trial tomorrow. So, we know we have the money. We got to go spend it and and then we're going to build a big business. Fantastic. Well, we're rooting for you. Amazing, guys. Congratulations. Good luck.

Thank you. Good luck, my man. Good stuff. Yeah, we'll see you. Fantastic. All right. These party poppers are a big problem. They like it basically was spraying like, you know, little shards of of of paper everywhere. But, uh, we're doing it. We're not going to stop.

I think it as long as it didn't go in my Red Bull, I think we're good. Um, let's uh let's ideally bring in the next person. You want to come on? Let's