Venture with Grace
00:05:20 · Vitaly Gordon
[00:05:20] S1: Yeah. So, well, starting with, you know, billing Einstein from scratch. So Salesforce, obviously, like, you know, like anyone who builds, even if it's a brand new kind of offering in a large company, you are, you know, somewhat standing on the shoulders of giants. Right? You are... You know, you have access to so many things. And, you know, in my case, was access to Salesforce customers. And so, you know, if we build a product that they want, then, you know, there's a whole machine that, you know, will make sure that they buy it. And so that that part was definitely easier. And I've, you know, throwing it all away and starting from scratch is probably, you know, one of the most humbling experiences that anyone can have. Like, I was a big shot at Salesforce, and then, you know, like, I was a nobody. Like, that's really what this is about. And you just have to, scratch and, you know, kind of claw and kind of just... Like, borrow, and steal in order to, get your first customer. There's just no way there's just no way around it. So the way I did it is just the old fashioned way. Throughout my journey at Salesforce and other companies, I, you know, developed some network and I was just like, you know, every day, I don't remember. I think I... And I was like, literally reported that in my monthly investor update, I put a goal and, like, how many people will I talk to this month? And I would just... Like, every day, would send 20 emails and I would personalize those emails and I would, like, try to say, hey. You know, I was willing to talk to anyone who was willing to listen. And then kind of like doing all this, you know, discovery was just kind of painful. But I... Like, you know, there's just no other way because quite honestly, everyone was busy and they had their own life that they had to worry about. And you just need to put in the effort, like to get some of their time to get some feedback and just go from there.
00:07:23 · Grace Gong
[00:07:23] S0: Since like, I feel like this year has been like people have been like either token maxing or agent maxing. But nowadays people are a lot more conscious about how much token they're using. So can we talk about how do you guys monitor the productivity of whether it's token usage or guess, like, the engineering productivity in general?
00:07:46 · Vitaly Gordon
[00:07:46] S1: Yeah. So look. This is something that obviously has been the journey since I started this company seven years ago where, like, it was pre GitHub Copilot and Codex and, you know, Claude Code and all that stuff. So improving engineering productivity was kind of like a relatively slow and laborious process that required a lot of change management. And now AI changed all of it. Like, honestly, there is no engineering productivity today that does not include AI. It's just like it's primarily, kind of how to use AI. And... But the good thing is, like, the results are now, like, massively, massively different, and we truly see, you know, like, these, you know, kind of proverbial 10x engineers that are, like, you know, orchestrating agents and outputting, like, so much more. Now, you know, we can talk about the quality of that. But, like... But that aside is, like, there's no doubt about that this technology is probably the most revolutionary technology that we ever had for in kind of engineering. And now having said that, I I think that some people are still, like, you know, just not, kind of thinking about it the right way because in the vast, vast majority of the companies, their AI spend is still less than 5% their labor spend, like the salary cost and kind of all of that stuff. I think, you know, Anthropic and S1 just reported that since, you know, like the beginning of the year, actually, their cloud code spend is more than their salary spend. So if you take a page from Anthropic, like, you know, like, again, as a percentage of, you know, engineering spend, maybe your AI spend should be over a 100%. Right? That's gonna look at least... That's what Anthropic can report. And so, like, there's a lot of talk about, like, you know, token maximum, which is also, you know, incorrect, because just like, you know, spending any kind of tokens in AI really does not yield a result. So the thing that we are kind of advocating for is outcome axing, which is really the thing that probably most organizations should be optimizing for, which AI is amazing at, is really lowering the price per outcome. Right? And by lowering the price per outcome, more often than not actually involves using more AI and not less AI.
00:09:57 · Grace Gong
[00:09:57] S0: You guys have... Like, so basically, you fund the engineering team with, like, the best pre AI practices, strong dura metrics. So I guess for a startup, how do you evaluate if your token spend or if you are spending the things into the right things. Then for the engineers, so I guess, like, how do you measure each person's like productivity? Because some of them are like defense, some are office slash, like, you know, people are, some are like writing code, some are like kind of evaluating the stuff. And like, nowadays, like people are also using different toolings on like different things. Like how do you picture where does your company sit in this kind of like, or like how do you guys evaluate like if something is more effective if you just bought in versus like buying the support?
00:10:48 · Vitaly Gordon
[00:10:48] S1: Yeah, look, so it's a, you know, it's a really hard time right now. Because here's the thing, the baseline just have moved so far up, you know, compared to before. So while, you know, on one hand, we can, you know, sit here and argue and, like, you know, taste is, you know, more important and, like, you... Which is true, by the way, but you still need to execute to kind of, like, at a level that is just will make, you know, like, what engineering team could produce five years ago seems laugh... Laughable. Right? So there is, like, if you're not producing, at a certain level, now we can, you know, debate and kind of industry by industry dependent, and we can, you know, share some of these numbers. But your competitors are probably, like, out executing you. That's really kind of, like, at least the way we go about thinking. And so if you're not also executing, so, you know, there's a thinking about ways, like, what used to take a year of development should now take two months. Right? Like, that could be, like, one kind of rule of thumb if, like, if you wanna be, you know, five times five times faster. Now it doesn't mean that, like, you should just, like, create, you know, massive, massive amounts of SLAB because it's actually pretty easy. You still need a, you know, you you still need a product that people will buy and solve, you know, customer problem and, you know, all of that stuff. However, you also need to kind of, you know, move just much faster. And so if you're not not feeling that, like... So I'll just give you kind of a rule of thumb. I think that if you still have, let's say, a ratio for, like, one one PM to, eight engineers, like... And the PM is not running out of ideas and because engineering is executing so fast, you're probably doing something wrong.
00:12:34 · Grace Gong
[00:12:34] S0: There's a term called code churn. Basically, it's code that's being written and deleted. In So your data, it's up to like 861%. And how is that kind of come together? How do you evaluate on like, how do people basically prevent this from happening or out of