The Exchanges

Q You got deleted. So, so for all seriousness, we've been doing this for a while. You built our whole AI VP of, of customer success. We built the saster annual.com together and Replit. What happened? Why were you deleted by your, your affirmatively intentionally deleted by the agent, your session. So what happened?

A Yeah, it's free sound. So we have an agent that powers our agenda, as Jason explained, and what happened was he just decided to be lazy because we started adding more and more sessions, right? We're closer to SASTR annual. We have more speakers coming in, confirming for SASTR in May. We had more speakers, so we added about like, 20 more just in like the last week. And the agent was lazy because he didn't update That he should look for all sessions that from the API, he just decided he would just stick to the core set that I gave him originally. And so when I added these new sessions, he, he decided he would just pull the first 50 instead of all of them. And so I got dropped, a bunch of others got dropped too, but this one was the most like visceral one because I just got dropped. But what was worse was the agent was lazy about it. He decided to blame the API integration. For why I got dropped.

AI assessment note: “he decided he would just pull the first 50 instead of all of them”

Q Just two questions. One tactical one on the data. So it cues them up before it sends in the beginning, what percent and how many messages did you read manually before you let the AI do it mostly automated?

A In the beginning, I read all the messages. So it was, you know, hundreds and then thousands. Like, I wouldn't start with a huge list the first time you do this. I think I started with a list of like, 600. That's a good mix because it's not going to find everyone. The match rate is fairly high, but it's not going to find everyone. So by the time it does that and identifies the right people, yeah, I did about 600. And then in the beginning, I would read because you can set the agent to say, how many messages do I want to do? So for us, historically, we always knew, like, if we wanted to email somebody to come to Sastr, An email from me or a human SDR would take about three actual human touches. And so we duplicated that in art and artisan and said, okay, three touches in email and then one LinkedIn message to the person. And so your outbox would then have four messages per person. So 600 times four, you're looking at 1200 messages. I would read all of them in the early days. Now I just spot check.

AI assessment note: “In the beginning, I read all the messages. So it was, you know, hundreds”

Q it in about an hour. So be wary. Is the SaaS apocalypse going to kill the leaders in B to B that are public? No, there's too much workflow. They're too rich. They're too complicated, but these point solutions like this B newsletter builder, we don't need it. We just don't need it. We don't need it. This is just better, right? So any thoughts to add to that, Amelia?

A It's, I think the only meta that I had was like, it's just our agents that started simple, even Tanki started simple, Kiwi started simple. Sastrainhole.com is now an agent. That was a website. They just keep getting better. The more you, you use them, the more you put into them, the more you think through, okay, actually an agent could do this too. They just keep getting better. I don't, there, there are times they do some odd things, but for the most part, they just keep getting better. They don't really regress in their career path as an agent and coworker of ours, and even as a potential manager of some humans upcoming. But yeah, it's just interesting. I think that I've gotten to this point where there are so many little, again, yeah, n equals one point solutions that they can just automate in a second. Again, this is like when I Get home super late right now during Loaded. I'm like, I don't want to think about the newsletter together. In prior years, I used to miss some, just straight up. I'd be like, you know what? I know we have sponsors.

AI assessment note: “so many little, again, yeah, n equals one point solutions that they can just automate”

Q That's the thing. That's why CS people fail. They gotta know the product hold, right?

A That's the, and that's the difference between as a CS person, a CSE and an FDE light. Like an FDE light, if you're trying to solve this problem, as long as they know the product, whether they come from sales, product marketing, product marketing, doesn't matter. If you have people in your org who know the product, they can be an FBE. They do not need to be an engineer as long as they know the product and can get your customer deployed. I feel like we're also losing the plot. That the whole point of the FDE is to get your customer deployed with your agent. So no matter how you solve that, again, that solution is not self-serve, but however you solve that with a person to get them into production and deployment, that's what you should have as your FDEs. Whether they're an engineer or not, I think by trade, I don't think it matters as long as I know the product.

AI assessment note: “If you have people in your org who know the product, they can be an FBE.”

Q And there is a chance some of this starts to come in, in, in, in cloud four, seven. But, um, but this is why now Figma feels like grandpa software. That's the thing. It feels like grandpa software. It feels like grandpa software and you don't, the NRR is still high. The company is still growing at top decile rates, but who wants to buy new stuff from grandpa software?

A Well, it's different too, but because Illustrator has slightly better agentic capabilities actually than Figma, like right now, so. This morning somebody said, hey, I don't like my booth graphics, which just drove me crazy. Can you fix it for me? And first I was like, no, I don't really want to, I told you, I was like, I don't want to take responsibility for your booth graphics. And they're like, I really just need to move the text and my designers out of office. And I know this is due Friday. So I, because it was an illustrator file, I put in an illustrator. I literally just asked the illustrator agent and said, can you change this thing? And it, and it moved it.

AI assessment note: “Illustrator has slightly better agentic capabilities actually than Figma, like right now”

Q How do you do both at the same time? Is it two different instances, two different orgs, two different iterations? And so, like, how many variants of this AISDR are, are, are multitasking, are running in parallel?

A Yeah, it's a good question. I would say in, in the first iteration, right, because this was an experiment for us, I mean, I know we're all, you're all in on I am. I'm all in. We all agreed we should try this. Um, before this webinar, listen, there's not a lot of content out there on how to do this for real. So I just started it. I was like, okay, let me use the marketing principles and sales principles I know since the dawn of time, but also let me just try different things. So I started it on three different experiments, one for each bucket, one, which was a totally cold outbound, literally to a list of people. Um, it was a mix. Some people knew Sastr, some people did not know Sastr because I wanted to see how the AI would do. Then I did, um, lapsed accounts, right? So these are folks that have known Sastr, but maybe haven't engaged with us in a while for sponsorships. And then I did one for tickets of people who have been to a previous Sastr event, haven't been to one in a while, see if that would convert. And then fourth was I had the website one on In a version that I don't have on anymore, but it was like basically doing some light qualifications of folks on the website, um, to see what would happen with the AI.

AI assessment note: “So I started it on three different experiments, one for each bucket”

Q But it's still much better than tracking, you spending all day tracking down people across the campus that never answer the question, right? It's so much better.

A And he's found, he's spotted not big issues, but he's spotted things that could have been an issue. Like this thing with the furniture order. 20 chairs were short in one area, but then over in another area. That affects budget. That affects the order. Then they have to go into the, that affects inventory. There's a lot of events going on right now. Like Anthropix having an event today in the Bay Area. There's a lot of events going on. That affects the truck. What if you're, what if all of those were in one area and you had no chairs for a whole section, then Those people are just standing or left there. Again, not things that maybe seem minor, but he's able to catch it because he can just ingest it instantly and give us an answer.

AI assessment note: “he's able to catch it because he can just ingest it instantly”