Jesse(00:00):
I don't think that there's any real secret. You just have to execute well and executing well entails building the right product, building it quickly, building it well, and then moving really aggressively on your go to markets and figuring out what is the next thing to
Sara (00:11):
Build. Welcome to Spotlight on a podcast about how companies are built from the people doing the building one messy, exhilarating decision at a time.
Sara (00:20):
Welcome to Spotlight On, I'm your host Ivan Zou with Decone CEO and co-founder Jesse Zang.
Jesse(00:25):
Thanks for having me.
Ivan(00:26):
Awesome. Would you like to tell the story of why I am wearing a hat when I normally don't wear hats for podcasts or really anything?
Jesse(00:35):
Yeah, I mean, so Ivan and I have known each other for a while. He's always had the same exact haircut, which is a semi buzz cut with longer on the top. Last year, he bet us that we couldn't achieve a milestone revenue milestone at the end of the year. He was just joking around. We had an amazing year last year, so luckily we did blow through it and then we forced him to shave his head. So he came to the office and we took some clippers and buzzed it all off
Ivan(01:03):
And it was all in front of the new team, the whole team too. So that was fun. Well, maybe tell us a little bit about yourself, your background and how you got to be the founder.
Jesse(01:11):
Yeah, so I am one of the founders here at Decagon. Before this, originally born and raised in Boulder. Grew up doing a lot of math contests, stuff like that. Led me to study CS at Harvard. Started a company afterwards that was eventually acquired by Niantic where we worked together. And then after that left to start, so the main motivation was my first company was a consumer app. A lot of customer support that was just done manually by me. I had to basically build all the systems. Back then there weren't really gen AI models, and so there was more old school automation. Through that process, I think I built a lot of empathy for both the support operation but also the tooling that's needed to make successful. And so at Ethicon what we do is we started, after GPD four came around, you could really see the potential of lms and so simply put, we're a AI customer service agent. The whole vision is that in the future, any user of any product, any time is going to have a personal concierge there that they can talk to, a response instantly has access to all their data. It can take actions for them and it can just solve any issue for them. So that's the goal and it's been a fun ride so far. I think it's definitely been one of the most exciting use cases for lms.
Ivan(02:32):
Awesome. So we first met each other while we're both founders and you were working on a company in a similar space, consumer gaming. Maybe you can share a little bit about the experience building low key and some of the learnings that you're bringing with you as you work on Tecom.
Jesse(02:51):
Of course. Yeah. I mean we obviously shared a lot of elements in our initial companies, obviously great chatting with you about some of the problems that we were solving. So I think biggest learnings, well first of all, that company was a consumer company. It is very, a trope that I now often see in hindsight is that younger founders or a new grads are doing consumer companies and same as me, you feel like you have to do a consumer company because you never worked anywhere. So how can you do a B2B company? Consumer companies are just very different. It's like you just have a ton of users. You don't really care about revenue for a while, and you're just worried about growth when you're thinking about how to grow. It's not really like you talk to customers and you figure it out. You can talk to some users, but you're only going to get so far from there. You're kind of just thinking through, it's very intuition heavy, it's very product heavy. And when we were building that company, it was during Covid, it was a remote company. We did a lot of just overthinking for sure,
Jesse(04:02):
Oh, this thing seems to have worked for someone else and we should really apply that somehow. And
Ivan(04:11):
A lot of reasoning by analogy or things that were not as ground truth
Jesse(04:15):
Exactly, or you see this trend somewhere
Ivan(04:18):
And it's like,
Jesse(04:19):
Oh,
Ivan(04:20):
It's so hard in consumer because what is ground truth in such a fluid and taste driven space?
Jesse(04:27):
So you end up just really putting a lot of brain cycles over analyzing this stuff, okay, how do we grow? And I think that's very difficult on consumer. I think a lot of the best consumer founders are just very good at intuition. They're not really thinking about stuff. I think a lot of the super quantitative, let's say people that did math contest, it's like those people I think generally are a better fit for more analytical projects where it's more you have customers or it's a deep problem or whatever. I think trying to analyze viral growth and stuff like that is very difficult and some people I think are just very good at intuiting through that.
Ivan(05:17):
We worked together at Niantic. I was previously a founder that sold my company. And then you had joined as well through the acquisition of Loki and I remember did you last a year at night and tick? Was it a year,
Jesse(05:29):
A year and a half. Around
Ivan(05:31):
A year and a half. Okay. I still remember when you pulled me into the room for our one-on-one and you told me, Ivan, I have to leave right now to start another company. And I did what anyone would do at the time and be like, maybe another six months, maybe we can get to the end of this project. And you were just like, no, no. The urgency was so strong. I had love to maybe just hear from you. I didn't get to ask at the time, what gave you that much conviction and urgency to leave at that moment versus staying a little bit longer?
Jesse(06:03):
Niantic was quite a nice experience. It was a great team, a lot of fun projects. I think for me it was one, the new technologies that were coming up, LMS obviously being the main one, were just super exciting and just felt like, Hey, in my twenties I really need to use this time to take big swings. And so that was the main motivator and really wanted to start something new. I mean, the first company wasn't necessarily super enjoyable the whole time. We were struggling for a bit, figuring out what idea to work on. Companies are never easy. It's a lot of stress. But I think even with that, we had a nice outcome. It was like at a time where I still felt like I had a lot of energy and so I wanted to take another stab and try something, building something new with all this fun stuff that was happening.
Ivan(07:02):
That's awesome. I remember one thing that stood out to me too was you just said that this moment in AI was the perfect moment for you in particular. I don't know if you remember that, but that was something that got me really excited at least.
Jesse(07:22):
So I think for me, what excites me and what I think I'm good at is really building around users and customers and the newer developments were all this new. We started basically after GP four came out, it was just so clear that there would be so many cool applications that people could use. And that was exciting to me. My background obviously was very AI ml, heavy math. Yeah, I mean it was just too exciting to ignore. I think the other way you could think about it is you want to minimize the regret in your life. So I would definitely regret not using these years to try something. And in hindsight, obviously it was turned out to be a good decision. But yeah, sometimes you just have to do what you feel like has to be done.
Ivan(08:16):
Awesome. Okay, so you left, you started exploring a few different ideas with early customers. I think something you guys did really well was customer discovery. I think that process is for many founders, but even companies at scale that are trying to build product number two or product number three. I think customer discovery is both an art and a science. And I'm curious, maybe you can share a little bit more about your framework, your approach. How did you go about navigating the idea maze with those early customers? And for context support wasn't your first, you didn't off the bat think support was the perfect use case. Nail it one shot. Right.
Jesse(08:56):
When we first started talking to customers, it was we tried to keep the mindset of keep an open mind, don't overthink things. You really want to think things from first principles. And because of that you want to block out the noise, block out what people are telling you, block out even what investors are telling you, especially what investors are
Speaker 4 (09:12):
Telling especially.
Jesse(09:13):
And from that you just listened to the customer. And the main sort of way we're thinking about it is you come into a call, you have some hypotheses around what could be useful for them, but you let them drive the conversation and then through the conversation you really want to get at how valuable is this actually? And that's the general framework we went through. And one interesting question would be like, Hey, if we built this, how much would you pay for it? It's a good question to ask because it separates a lot of the signal out. And someone might have a really excited conversation with you maybe because maybe they owe you something insightful because you're on a call with them or something, whatever the reason. And at the end of the day, it's like, oh yeah, actually no budget's tight. Maybe like 50 bucks a month, stuff like that.
Jesse(10:10):
So that helps you separate the signal. And so at the beginning we were just talking about all sorts of ideas. It's like anything that could potentially be helpful for people or save them time or have an agent do work for you. And so a bunch of ops use cases, data use cases, pre-sales, security, literally everything, because at the beginning it's very hard to pigeonhole yourself into something and so you just want to talk to all the people that will be down to talk to you. And yeah, I mean I think that navigating that path, there's obviously a bit of luck involved for sure, talk to the right people, you have to be at the right time. But people actually very quickly pointed us towards this direction. We talked to. I think the general progression was we talked through a lot of ops use cases, and when you get to the end of the conversation, you talk through this whole thing that you could automate or you could have agent do, and you're like, okay, how valuable is this to you or how much would you pay for it? And then they start thinking through like, oh, well I have maybe three full-time people doing this, and if you do this really well, I could have maybe one of them go do something else. It's not, the numbers aren't going to be that big. And then a lot of people proactively told us, oh, but by the way, we have 500 people in our support org.
Jesse(11:31):
If you could do something there, it'd be really valuable. And then you just go down the normal discovery path. They're like, okay, yeah, sounds good. What are they doing now? Tools are they using? What are the issues? And it's like, okay, would this be a good solution to your issue? What about this? It's like, okay. And then when you asked the question at the end, how valuable is this to you? It was very valuable. So I think that's the general way to do it. And it's not always going to work, right? Because there's a lot of other factors. You have to have consistency over the use case. If something's valuable just to one company and you're basically building software for that one company,
Ivan(12:10):
It's more of a service.
Jesse(12:11):
Yeah, it is valuable, but it's not really a software company. Totally. You can't just ask that question, solve it. But yeah, we explored all sorts of different things and through that process found this idea.
Ivan(12:28):
Awesome. I remember in the early days, the market does seem crowded from the outside, and this was early last year, but when you ask customers, what do you use? None of them were really using true native generative AI in production and they had maybe demos in place. So maybe you can talk a little bit about that leap from an early demo with those first, let's say 10 customers to something in production. What did you have to build maybe the engineering and software around the LLM to actually make that work in production in a way that the first version that you'd built?
Jesse(13:10):
I think in a lot of situations, if someone was using an incumbent software or older versions of a product and then they did have a good gen AI offering or something, then maybe the conversation wouldn't even have gotten to that point because then they would just be like, okay, well yeah, I mean the current thing is pretty good. When we first started, that was definitely not the case. And it was a little bit surprising to us because again, it does make sense for a lot of, if you just think about the previous version of what we're doing, which is maybe like a chat bot or IVR, those sort of things,
Ivan(13:42):
Like a decision tree.
Jesse(13:44):
You go on a phone call and it's like, press one for this. It's very natural for those people to have the next step be gen ai.
Jesse(13:53):
But for whatever reason, they moved relatively slowly and the quality was not very good. And so when we first started, there was actually a really big difference, and it goes to show that when there's new technologies, you really have to put aside, again, thinking from first principles, you have put aside what's already been built and what your infrastructure already is, and just think about, okay, now that there's ion models, what is the ideal experience if we had to build from scratch? And that helped us. So that got us in the door and going from a demo to a real product, it's all about pace, I think. So most interesting spaces or most hot spaces, there'll be a lot of players, a lot of people that are interested in the space, and there's really no secret to it. People can copy each other's products, people can say the same thing as each other.
Ivan(14:45):
Yeah, all the marketing sites start to,
Jesse(14:47):
Yeah, exactly, start to sound the same. I don't think there's any real secret. You just have to execute well and executing well entails building the right product, building it quickly, building it well, and then moving really aggressively on your go-to markets and figuring out what is the next thing to build and things like that. And in the past year, I think we did a good job of that, and so we're really trying to make sure we keep that up in the next
Ivan(15:09):
Year. Got it. Awesome. Maybe transitioning to the product a little bit and going deeper there, you ended up focusing on a vertical or a use case I should say, with support. And there's been a lot of talk in agents generally around general purpose employees where you hire an agent and you can deploy it to different use cases versus focusing on one specific use case and maybe taking it one at a time. And we've talked about this in the past, but how do you think about those approaches, building kind of more general purpose agent versus just taking one use case and really dominating that?
Jesse(15:50):
I mean, I think it's hard to have a real agent that reaches his full potential without actually going deep in a use case. The main value for a lot of what we're doing and maybe what other agents are doing, it's like when we talk to customers, we weren't really thinking about it as an agent. We're thinking about it, okay, what is the problem that you guys have and how can you solve it?
Jesse(16:12):
And the problems tend to be vertical specific, and so if you need to fully solve the problem, you need to, you're basically just building software at that point, right? You're building all the tools for that team to manage the agent, to audit it, to get value out of it. In our case, our agent has a lot of conversations, so you need to be able to deploy the conversations in a way that reaches their customers. You have to integrate with their systems. I think most of what we're building is honestly just software. It's application software that fits this use case. Of course, what the software ends up doing is we're bringing an agent to you, but it's still there to solve that team's problem, and that team doesn't care about other use cases, they just care about their use case. So I think you have to go pretty deep when you're at the application layer,
Jesse(16:58):
Otherwise the amount of value you're providing is not going to be enough for a team or a company to justify a large investment. Most of the difficult part is like, okay, the AI models are here, they can do a lot of cool stuff. How can you mold them around the customer's business logic and what they want? And not only that, okay, so once you do that, you maybe have an agent that kind of works, but then you have to think about all the tooling around it for their employees to audit the conversations, to QA things, to build logic into the ai, and that's where most of the work goes towards.
Ivan(17:32):
So you guys have seen really rapid adoption over the last year. How have customers been responding and reacting? Obviously they're using the product in production, but would love to hear more nuanced takes from you on what you're seeing from customers, especially given the newness of the technology and how you're almost building it as you're flying the plane in some sort of form.
Jesse(18:00):
In our space in particular, customer service, customer experience interactions, like no one really needs to be taught about the space and no one needs to be taught about what AI agents can do. So that is actually a really nice property about the space. We don't have to do the education and convince people, Hey, you should do this. It's more that people are, AI might be a new thing. So in their mind they're like, okay, let's slow down a bit. Let's fully understand how it works. There's also the sort of black box issue where if you use chat GBT, the answers are awesome, but how did it actually come up with it? Who knows? And at least that's from the customer's perspective. And so they're a little bit scared to really dive head first into something if they don't understand how it works, they don't feel like they have control over it. They don't feel like, Hey, here's a conversation that happened. How exactly did AI come up with these answers and what the steps it took? And if I want to change something, how do I do it? So that's helpful for us. I think we use that to motivate a lot of the product that we built, which is really building out this tooling to make it not feel like a black box and give people the guardrails and give people the keys to drive.
Jesse(19:16):
So I think that's generally the reception. Of course, after someone uses it, the results are very clear. And so we have a bunch of great customers, a bunch of great testimonials. People are seeing really tangible results. And so that's all great, and you really need that in a space where you're building AI agents. If you didn't have the results, then it's way harder to get adoption. So now you think that's where we're at right now, which is we have a lot of customers, obviously we're focused on getting new customers. A lot of our current customers, I'll probably describe as early movers or early adopters, like the people or they're large organizations, but they're very tech forward. They're very sophisticated. They know how to evaluate things. I think this year, this space is getting towards, I guess the second portion of the bell curve, which is the early majority or whatever it's called, where you have a lot of bigger companies that now they're really bought into Gen ai.
Jesse(20:15):
Not only are they bought into Gen ai, but they know that the most obvious use case for them is customer support, customer service. And so yeah, they already bought in the use case, so they just need to be convinced that like, Hey, this is the right way to do it. There's a bunch of different ways that customer support changes. The obvious one is that it becomes a lot more efficient. So the people that are working in customer service, a lot of the mundane tasks just go away because they're done by an agent. And if you had a choice, of course you would choose for it to be done by agent, because for the customer, it's better. You just get it instantly. It's super consistent for the company. You don't have to like, okay, a new change happens. You have to retrain all your agents, you have to worry about attrition, you have to worry about inconsistencies clearly better for both sides.
Jesse(21:05):
And so now what the human agents are doing are more of the complex tasks, the higher order tasks, it's more cognitive, it's more, I guess intellectual, I suppose. So I think the nature of the job changes, that's one big difference. The other one is I would say that people start thinking about their customer experience org differently because in the past, what ends up happening is that you might just see your customer support or your call center or whatever as a call center. It's like it's probably not core to your vision, it's probably not core to your annual goals, but it's there. It probably costs a lot of money. So maybe the only goal you have is reduce the amount of money, but with agents, they're so customizable and they're so powerful that it starts becoming competitive advantage. And to an earlier point, it starts becoming a revenue driver. So now all of a sudden C-Suites really care about it. The CIO really cares about it. It actually is a thing that can push the company forward, increase your revenue, increase your customer happiness, and it can be really core to your product as well. It can be a core part of your product because maybe your product's very complex and if you have conversational entry point, that's great. So I think that's a big shift. It's one the shift of the work that the support org does, but also it shifts how leadership thinks about support and customer service into a much more sort of aggressive forward thinking initiative.
Ivan(22:38):
Awesome. Well, maybe we can also transition a little bit to how you've built out the team. I think when we rewind to earlier last year, it was just the two of you all the way until maybe February, March, and you've, since I think the team is 30 people or more now, what are some lessons that you've learned from hiring and scaling the team so quickly?
Jesse(23:01):
Asha and I would say have always been pretty conservative about hiring because the leaning has always been towards a smaller but super stacked team. Rather than just scaling headcount and doing large scale headcount planning. I mean, at a certain point you do have to do planning, and we do planning now, but we really just wanted to find the best people that could really build a culture that works. People get along, people are working hard, but it feels like the work is meaningful. So that's the approach we had. I think up until the beginning of this year, it was just hire as quickly as possible, but keep the bar really high. If we found a really good person, we would be able to have meaningful work for them. So building out the team currently, our team is roughly the three categories are engineering, e, PT, basically, and then go to market and ops. And ops for us is making sure deployments go well and things like that.
Ivan(24:01):
Got it. Dipping a step back from hiring then. Are there other frameworks that you feel like you fall back on a lot as you make big decisions throughout building the company and including Ashwin as well, between you and Ashwin? When it comes down to, I remember actually one interesting, a discussion we had early on, was it important to own the system of record for agents? And we could maybe use that as an example, but yeah, I'm curious if you have some frameworks and you can maybe couch them in one or two big decisions you've applied them in.
Jesse(24:36):
I see, okay. You're talking about more product frameworks,
Ivan(24:38):
Product and specifically for in this new standard of
Jesse(24:45):
Software. So ground truth is always making sure you're doing what customers actually care about. And that goes into the beginning of how we even start working on this,
Jesse(24:57):
But we spend a lot of time with customers. We really try our best to prioritize what we hear from customers. And it's not always like, oh, someone says this feature, you have to build this feature. It's more like, we shouldn't just think about things in a vacuum ever. Everything should be motivated by what you're seeing from a customer. It's not that the customers will tell you everything, but at least you can read between the lines sometimes and based on what their initiatives are and so on, that will be the best signal for what to build.
Jesse(25:27):
And so when we think about that, that's mostly the motivation for everything that we do. One big theme that we've thought about recently is in that vein, using that framework, it's like, what do customers really care about? We've realized that when people think about AI agents, there's a lot of value and expectation they want to put on it. The easiest one is, okay, adding more efficiency and saving costs, that's the easiest one, and everyone gets that immediately. But what really gets people super excited is if you can extrapolate that to, okay, now your customer experience is a lot better
Jesse(26:04):
And that actually leads to more revenue or more bookings or more upgrades. And so I think that's an example where you have the framework of, okay, customers care about this. You heard that from people, and how do you apply that to the product? You want to be more proactive. You want to be able to allow customers to build that into their agent. The whole point is where is your North Star? So I was just saying our North Star is customer feedback, but another North Star could be you want your product to be as sticky as possible,
Jesse(26:34):
And for some people that means they're a system of record. So system of record just means that you are the store of the ground truth of some sort of information. In our case, a Zendesk could be a system of record. So Zendesk, what they do is they're a ticketing system. And so when people, if they've been around forever, so if any product you had an issue, you run into their email, that email probably goes into a ticket in Zendesk, and now over the years, a bunch of tickets get collected in Zendesk. So that's like a system of record. Salesforce also a system record, and system records tend to be really sticky. That's why Salesforce is so sticky, and it's such a cool product because even if you don't like the product, you just have to keep using it. And I think that's a great spot to be in. I mean, ideally people have to keep using it and they like your product, but system record basically makes it super sticky. There are other ways to make things sticky. I forget the system of whatever, but there's a lot of systems, system of actions, system of actions where, okay, another example is your product is sticky because people keep using your product every day and their whole team every day. Their workflow is they log into your product and they do something.
Jesse(27:47):
So if that's the case, you also become sticky. And so I think these frameworks are helpful to maybe use as North Stars when you're planning out product. Again, I think going back to my initial point, we try not to think about these too much. You can kind of fall into traps where you're like, oh, I think it's really important to be a system of record, and then you just start building stuff. Great point. Maybe people don't really like that or even want that, and so it's just kind of a waste of time. And also you just build something that people don't even
Speaker 4 (28:15):
Like.
Jesse(28:16):
But it is something to just think about as you're planning out like, oh, I'm gathering all this customer feedback. Does this help us become more sticky? That is important.
Speaker 5 (28:24):
Totally.
Jesse(28:24):
And if you're seeing problems of like, oh, actually your turn is really high because it's not sticky, maybe you should think about that more, but it's not something that motivates us too much yet because it feels like we have a torrential flow of customers asking us stuff or telling us stuff. And that's really where we should be mining our product decisions.
Ivan(28:47):
Awesome. So you guys have done an incredible job of getting into customers, getting time to value very quick for them. And part of that is actually just your use cases. So clean cut. Maybe you can talk a little bit about how you think selling software that fundamentally agents might be different from selling SaaS software and how that evolution kind of plays out.
Jesse(29:19):
Yeah, good question. So fundamental difference, at least that we've seen is that in our use case, the agent agentic use case, there's a very good benchmark for the value you're providing, which is labor essentially. So how many tickets can you resolve? What's the customer satisfaction on those? Because those are very measurable. When you talk to a customer, very clear the value they're getting, and because of that, it's easier for them to make a business case and deals move a bit faster because they can just say, Hey, this is really working. Not only are we saving a ton of money, but customers are happier. And then ideally, as I mentioned before, they can also map, okay, customers are happier and their bookings are going up or their upgrades are going up. So if that's happening, it's like a very slam dunk business case. I think any exec will understand that or any leader. And so because of that, you're able to move forward and then you can also quantify the value because you had a bunch of costs doing that before. I think that makes things move faster. I would also say it makes the deal sizes a bit larger too, because you're able to quantify its labor is almost always order of magnitude or two more in terms of cost than software.
Jesse(30:40):
So if you can tie it to labor and like, Hey, there's actually ROI here. There's not always a one-to-one linear relationship, but you can generally get larger contracts because of that.
Ivan(30:52):
Got it. Do you find that gets easier over time as you add more tooling and functionality and capabilities to the product, or do you find that during the sales process, the demo, the time to value has not changed that much because you're not deeply embedding into the workflow yet, you're just kind of getting something up and running quickly to show value?
Jesse(31:20):
Time to value in our space is pretty quick actually, because I mean, this is a nice property of customer support is that you can just start with one use case and get it deployed and you can escalate the rest of the conversations still. And even if you just start, let's say you're just doing something super rudimentary and you're solving one third of the conversations, that's still a lot of value for them. And I think a lot of other use cases don't have this property where you can just get started easily and incrementally build up. And even with the initial thing, you can deploy into production, whereas other use cases, it just might have to be near perfect before you can deploy it, and then the conversations become harder. So I think that's the nice part about this as well, where the time, the value is fast because you can just try it out and then after you try it out, you can see the business case.
Ivan(32:09):
What does adoption look like from customers right now for what you've already delivered in that
Jesse(32:14):
Direction? So what we've delivered right now, it's a combination of a bunch of things. So we have ways for customers to define logic. We have ways for them to test it. We have ways for them to, after it's in production, monitor it and QA, as you said. And I think people really like those features because it gives them more, basically, it doesn't require them to do that much work, the work's done by the agent and a lot of logics built out, but it gives them peace of mind and control and visibility. That's the transparency. That's a big piece of what's happening. And can I make changes? And I would say almost every team we talk to that is important. Even if you talk to a team, they're like, Hey, we're really strapping resources. We provide the help to spin up their logic and stuff like that. They still want the ability to view things and to audit things, and that ends up being very critical.
Ivan(33:12):
Awesome. Maybe we can just imagine forward a little bit around what you think agents as a paradigm will look like in the next few years. What are some key things that stand out to you?
Jesse(33:27):
I do think it depends on the space. I don't think all agents will be doing the same stuff. If you think about, well, I think the other type of agent that's exciting right now is coding agents.
Speaker 5 (33:39):
Yeah,
Jesse(33:40):
Coding agents are very different because you basically have to have someone reviewing them, but even if you don't, they can still do a lot of stuff. What is the paradigm? I think the measure for how good an agent is is always going to be benchmarked around labor. So is this doing real work that is valuable and can I tie it to work that our team otherwise would've to do? And so in our space, the work is talking to customers, maintaining that relationship, solving their issue. Our agent now does even more work after conversations. They'll review them, you can tag them, it can review them for missing information, write new articles for you, stuff like that. So that's all work that's being done. And in the future, I think you're always going to be benchmarked towards work. Right? And same thing with coding agents. You're kind of benchmarked to the work a software engineer would've to do So. Yeah, that's probably the best paradigm because usually with traditional software, that's not always the case. Maybe it's just the productivity tool, and it is very hard to tie that to work.
Ivan(34:51):
It's much more about amplification,
Jesse(34:53):
Amplification or just enabling something not been possible before. And now it's possible. I think agents, it's more about just doing
Ivan(34:59):
Work. Cool. Awesome. What do you think about the next year for deagon? What are you most focused on? What do you think you guys have to nail and maybe what would be the thing that keeps you up at night?
Jesse(35:15):
The thing that keeps me up at night is definitely just we've been growing really fast. You don't want that to slow. And so you want to look ahead, see what could possibly cause it to slow, and then make sure that you don't do that. So right now, we are focused on building out the team. We're focused on putting the right processes in place so that as we scale the team, things don't slip. Marketing's a big focus as well. So making sure in the next year that, Hey, we have all these great customers. Other people know about them. Other people know about the use cases and the implementations, and that if in the future people are considering this use case that they chat with us and we can see how we can help. Things like that.
Ivan(35:58):
Awesome. Cool. Well, any final things that I didn't ask that you would love to share with the world or share with me?
Jesse(36:06):
No, I mean, right now we're aggressively winning the team. We're really looking for people that are strong and excited about AI agents, so engineers, research engineers, infra engineers, product engineers. We're looking for product person as well, a designer. We're looking for all sorts of go-to-market people across the stack. So basically any role imaginable we we're looking for. And so yeah, there's people excited about AI agents and these use cases. Happy to speak with them.
Ivan(36:39):
Cool. Well, thanks for joining us today. Yeah, thanks for hosting me. Awesome.
.jpg)
