decagon

An AI-powered platform providing enterprise-grade generative AI agents for customer support, enhancing efficiency and customer satisfaction.

Decagon

Hey everyone. I'm Jesse. I'm the CEO and co-founder of Decagon. We build AI agents for customer support. 

So I'll start by just walking you through a little bit of our traction so far. We started the company about 7 months ago. Right now we're doing seven figures ARR, all from fairly large contracts from customers such as Eventbrite, Rippling, ClassPass, um, and so on. 

What specifically what we're doing for them is we've deployed these AI agents out. We're handling their global support volume and essentially just doing all the things that traditionally would have required a human agent to do. 

So if you've been following Geni for the last year or so, I think you may have. 

Please share the AI Agent Store with friends.

Noticed two things. One, people have tons of ideas for different AI agents. But overall, the sort of enterprise adoption for them hasn't been that great. And for the ones where there has been adoption, um, such as support, you, it kind of seems like all right, it's a super crowded space, right? Tons of people going after it. It's a very obvious idea. 

So, so today I'm basically just going to be talking about our progress and learning so far. How we've gotten these customers and tell you a little bit about the company. Start with the team. So this is me and my co-founder Ashwin. Uh, we have pretty similar backgrounds. We were both engineers. We started companies before and successfully sold.

Them and the whole vision here right is we are very excited about kind of realizing the vision for AI agents in the Enterprise and building a big company in the process. 

So back to specifically how are we getting these customers right there's a lot of players out there. Why are we winning all these deals? This is a nice quote from one of our customers uh substack and ultimately right now when you're selling to Enterprise it comes down to kind of capturing business logic the best. 

What I mean is every company is different they're G to have their own sort of nuance requirements and sort of customizations. But you kind of have to go really deep and be able to be flexible. 

Share the AI Agent Store with friends.

Around the requirements and the challenge is you can't build a custom product for every single company because that's not really scalable. So it's how can you build something that's scalable and flexible and really you know find the balance there. 

So what we've done is we've built our own orchestration layer above you know a variety of language models that can take you know questions from customers and do a good job of everything a human would do right looking up data doing actions and so on. 

So to give a very specific example substack, a user might come in and say something like uh hey I really want to cancel my subscription to John's newsletter. 

In reality they

Might not have uh a newsletter that's actually called J's newsletter. Uh it's just written by John. So the AI agent has to go through and knows which data to pull. He'll probably have to look at all their subscriptions, pull into authors, figure out like hey are you talking about this one. The user confirms and then actually goes and does the cancellation refund. Right, so that's a pretty simple example of like what an agent would have to do. 

And that's kind of the difference between AI agent and something like a traditional chatbot. Where even if it does have j in it, it would never be able to do that. So that's how we're winning right now. Right, just based on quality and just. 

Share the AI Agent Store with friends.

Quality of the product. The reason why the space is so popular and so exciting in the future is that the ultimate goal here is really being able to build software that can replace the sort of massive services spend uh that people that's going towards a human agents right now. 

And it's very hard to get there. Like human agents do more than respond to questions and even that is hard enough. But they'll have to read all the conversations. Um at the end they'll have to tag them. They'll figure out product insights, identify trends, figure out gaps in the knowledge base and fill it in. You know, file bug reports and feature requests. 

And so all of this stuff should be able to be shared with friends.

Done by AI agents and just right now it's it's not there yet. And so we want to lead the charge to be you know the first to unlock that value. 

So far um this is the status of the company. We've raised 37 million from Excel Andre Asar and Elad Gil. We're a pretty unique time right now where you know we haven't really officially launched. The team's still super small. We have a ton of resources. We have some good good early traction. 

So we're really aggressively looking for great people to join the team. So we're really looking for engineers, salespeople, chief of staff. On the engineering side it's kind of a fun time because there's like so much to do across the stack. 

Please share the AI Agent Store with friends.

Also a ton of really customer facing work where we're working with these huge C uh companies. Ton of requirements moving really fast every day. 

And the last detail here is that we are 100% in person in SF. So feel free to talk to me afterwards. I'll be around. I have this nice uh decagon sweater on and uh thank you all for coming. 

Tell about AI Agent Store to your friends.

Note, that there may be slight differences in text and video. Watch video for the original content.

Decagon Video