The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 33 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Zhang: Most AI agent use cases lack commercial viability with current models
“I think our view is that for the vast majority of use cases right now, it is still, like there's not going to be real commercial adoption with the state of the current models because of a bunch of things.”
Zhang: Enterprise adoption of cybersecurity AI agents will be slower than expected
“And so because the models are inherently non-deterministic, It's very hard for buyers like really trust a gen AI solution there. And so, especially agentic solution. So like, I think that option there is going to be really, really, really slow, a lot slower th…”
Zhang: Decagon cuts enterprise contact center costs by 60% to 70%
“And we've done case studies now where, you know, folks have been able to cut that down by, you know, 60, 70%.”
Zhang: Decagon prioritizes raw intelligence over direct experience across all roles
“We're generally just selecting for very smart people. First of all, I think we care more about that than like, you know, direct experience and so on.”
Zhang: Decagon mandates five days in-office, with many working weekends voluntarily
“We're five days and then a lot of folks come in on the weekends, but it's not like a requirement.”
Zhang: All successful AI companies share heavy in-office cultures
“I think pretty much all the AI companies that are doing well have, you know, pretty heavy in office cultures. It's just, you get way more done, especially, especially in the early stage.”
Zhang: Aspiring founders learn more at post-PMF startups than pre-PMF companies
“Generally, when I talk to engineers that want to join startups, for example, and let's say they eventually want to start their own company, which is a very common profile, it's, in my opinion, it's like way more useful to join somewhere where they've already k…”
Zhang: Foundation model labs will target consumer applications before enterprise software
“I think it makes a lot of sense for them to push into application layer, and I think they will. In terms of what applications, I mean, generally they'll probably start with applications where it's more consumer prosumer-y because there's, it's just more self-c…”
Zhang: AI labs will target coding before customer service
“I think before they tackle our space, there will probably be other spaces they have to tackle first. Yeah. Coding's probably one of them.”
Zhang: Salesforce's configuration-heavy SaaS model is wrong for the AI era
“Again, obviously, we respect like the sales forces of the world that build amazing businesses, but we just don't think that's the right approach for the AI era.”
Zhang: AI agent industry will shift toward output-based pricing
“And we'll probably start seeing that more and more in the AI agent space where you generally price per like the output that that's doing. I think that that works. I think that's just very clearly the right pricing model for our space”
Zhang: Consumers Will Interact With Agents Over Apps and Websites
“And eventually, if it's good enough, most consumers will just interact with the agent instead of even logging into the mobile app or the website.”
Decagon AI saved Bilt Rewards 65 support agent headcounts in one year
“Now basically we're almost, almost a year in at this point, they've been able to really restructure their customer support team. And again, we published a case study on this where they were able to quantify like, okay, what are the savings? Right. And so, so f…”
Zhang: Instruction following matters more than reasoning for customer service AI
“And for us, actually those things help, but they're actually not the biggest difference maker. So in our use case, the type of intelligence that matters the most We would probably describe it as instruction following.”
Zhang: Viable AI agent use cases require incremental value and quantifiable ROI
“The use cases that emerged, like you have to have those two qualities. Like it has to be able to be something that can be rolled out slowly and doesn't have the perfect off the bat, but it's already providing value, right? Like I think coding agents is like a …”
Zhang: Enterprise AI adoption is top-down, with customer service as lowest-hanging fruit
“Yeah, I mean, another specific dynamic is that at the enterprise, it's becoming a lot more of a top-down motion, so You know, in the past, any of these technologies could have been just, like, one team trying to vet it or decide it, but now it's like a, it's a…”
Zhang: Technical Founders Have Untapped Potential When Embracing GTM Problems
“One of my I guess theses is that there is a lot of untapped potential and just like really strong technical folks in making them a bit more commercial because the types of problems on the go to market side there, I would say generally a little bit more hairy. …”
Zhang: Post-PMF founders must shift from short-term to long-term thinking
“I think at the beginning you have to short, you have to think short term. You're just optimizing for closing the deal or closing a couple customers, but once you have your legs under you both can think more long-term and also you have obligation to, because if…”
Zhang: Most OpenAI revenue and margin comes from ChatGPT, not API
“I think OpenAI, for example, most of their revenue and most of their margin for sure is coming from ChatGPT and the application layer, because you actually own the customer”
Zhang: Engineering teams want to offload AI agent logic to business users
“What we found is even when the engineering teams are very much involved, they don't necessarily want to be on the hook for every little change, and so in that case, we can work very well with them, and you have them still owning how does the AI agent interact …”
Zhang: Pricing AI agents per minute incentivizes needlessly long calls
“You also don't want to price per like, you know, minutes of the call either like that. That's just kind of weird. And also incentivizes the agents to just like have really long calls. So you price basically the number of conversations that it can have.”
Zhang: Near-Term Agent-to-Agent Communication Will Remain in Natural Language
“I think in the near term, they'll still communicate in natural language, just because, like, each agent also needs to be compatible with humans, right? So if they talk to a human agent, a human support agent, or if we talk to a human customer, of course it has…”
Zhang: Customer Support AI Will Expand into Upselling and Proactive Outreach
“Right now a lot of the conversations are more reactive support. It's like, hey, I have an issue. Can you fix it? But over time, it'll be more and more kind of broader, right, in terms of, like, being able to do purchasing decisions, being able to upsell folks,…”
Zhang: Customer service is the golden use case for AI agents
“Our current use case as maybe what we think is like the golden use case for these AI agents, which is customer interactions, customer service. The use case is very tailor made for what LLMs are good at.”
Zhang: Most AI application alpha lies in orchestration and surrounding software
“Most of the sort of alpha or most of the specials stuff that you build is on top of models. It's either the orchestration layer or the software around it.”
Zhang: End-to-end voice AI falls short when complex tool use is needed
“Voice-to-voice, latency is great. Sometimes, though, with these production use cases, you do need the extra computation cycles. So, you know, fetch data, do multiple model calls or there's, there might be other reasons that you can't do voice-to-voice.”
Zhang: Math Olympiad talent has shifted from quant trading to startups
“In the last few years, maybe it lasts like, you know, five, six years, like there's just, because startups have been a lot more mainstream, a lot of Folks in this demographic have gravitated towards startups as opposed to, you know, traditionally it'd be eithe…”
Zhang reveals reciprocal angel investing among Math Olympiad AI startup founders
“I angel invested in a lot of the companies you just listed. A lot of their founders are angel investors in our company.”
Zhang: Anthropic's computer use capability is not yet production-ready
“Like we've seen the computer use demo from Anthropic. Probably in my opinion, not production ready yet”
Zhang: Decagon is approaching 200 employees
“Never really needed that, but, you know, we're approaching 200 people.”
Zhang: Customer service AI drove six-figure demand at zero ARR
“The real answer is we just saw a lot of folks that were willing to pay us like, you know, six figure contracts, which at the time when you're at zero ARR, it's like,
oh wow, that's huge.
And a lot of folks that were willing to, you know, do the same thing.
And…”
Zhang: Consumer AI Agents Contacting Enterprise Support Not Yet Happening at Scale
“It's not something we're seeing at scale now where you have agents writing in for you.”
Zhang: Decagon is evolving into a conversational UI concierge for brands
“And as we, as we've grown, it's kind of becoming more and more of you gotta think of like a conversational UI for the brand where it's how every user can interact with it. And we often use the term like concierge to describe this, but that's what we do.”