Meeting Title: AI Report Discussion with ABC Date: 2026-01-01 Meeting participants: Zoran Selinger, Amber Lin
WEBVTT
1 00:05:02.990 ⇒ 00:05:04.210 Amber Lin: Hi, Zorong.
2 00:05:04.240 ⇒ 00:05:05.160 Zoran Selinger: Hi, Amber.
3 00:05:06.330 ⇒ 00:05:07.280 Zoran Selinger: How are you?
4 00:05:08.080 ⇒ 00:05:09.720 Amber Lin: I’m good, Happy New Year!
5 00:05:10.020 ⇒ 00:05:10.869 Amber Lin: What time is it?
6 00:05:10.870 ⇒ 00:05:11.270 Zoran Selinger: January.
7 00:05:12.170 ⇒ 00:05:15.609 Zoran Selinger: It’s… it’s 2.30 PM.
8 00:05:16.400 ⇒ 00:05:19.590 Amber Lin: Oh, okay, great. It’s not too late. I…
9 00:05:19.590 ⇒ 00:05:21.959 Zoran Selinger: No, no, it’s not, it’s not, this is nice.
10 00:05:22.340 ⇒ 00:05:29.430 Amber Lin: Yeah, it is, 5.30 a.m. for Robert. I actually don’t know.
11 00:05:29.600 ⇒ 00:05:30.679 Amber Lin: how he’s…
12 00:05:30.680 ⇒ 00:05:31.730 Zoran Selinger: Really?
13 00:05:31.730 ⇒ 00:05:35.579 Amber Lin: Yeah, he’s on the… he’s on the West Coast.
14 00:05:35.850 ⇒ 00:05:37.449 Amber Lin: So, he’s 3 hours…
15 00:05:37.450 ⇒ 00:05:39.969 Zoran Selinger: You see, usually you’re on the West Coast.
16 00:05:40.290 ⇒ 00:05:42.490 Amber Lin: No, he’s visiting family.
17 00:05:42.490 ⇒ 00:05:46.970 Zoran Selinger: Oh, okay, okay, okay, good, good. I think he might have…
18 00:05:47.130 ⇒ 00:05:50.620 Zoran Selinger: Misjudged, then, when this meeting should be?
19 00:05:51.870 ⇒ 00:05:56.000 Amber Lin: Yeah, let’s… I just pinged him. Let’s check. Okay, great.
20 00:06:16.330 ⇒ 00:06:18.260 Zoran Selinger: Maybe…
21 00:06:19.830 ⇒ 00:06:28.119 Zoran Selinger: give him another 10 minutes. Like, usually if I, if I’m waiting for someone, I usually give 10 or 15 minutes.
22 00:06:30.800 ⇒ 00:06:31.800 Amber Lin: Cool, okay.
23 00:06:31.800 ⇒ 00:06:36.649 Zoran Selinger: Yeah, so, I read the report.
24 00:06:38.070 ⇒ 00:06:39.389 Zoran Selinger: A few days ago.
25 00:06:40.060 ⇒ 00:06:44.270 Zoran Selinger: It was a really good one. I thought there might be,
26 00:06:44.530 ⇒ 00:06:50.680 Zoran Selinger: I didn’t really think that literally the whole thing is going to be about AI, but it’s…
27 00:06:50.680 ⇒ 00:06:51.340 Amber Lin: Yeah.
28 00:06:51.340 ⇒ 00:06:56.079 Zoran Selinger: It is, it is relevant, it is relevant.
29 00:06:56.460 ⇒ 00:07:01.129 Zoran Selinger: I don’t know if you, if you talked about it with, with Robert.
30 00:07:01.130 ⇒ 00:07:17.740 Amber Lin: Oh, not yet. I just read through it as well. I didn’t go through the interview section, but mostly I’ll just be providing some questions to guide the conversation. It mainly will be you and Robert talking about it.
31 00:07:17.740 ⇒ 00:07:21.350 Zoran Selinger: Is that, the notion, Doc?
32 00:07:21.560 ⇒ 00:07:22.100 Zoran Selinger: That I…
33 00:07:22.100 ⇒ 00:07:32.609 Amber Lin: Yeah, I tried to… So that’s yours? Yeah, but I wanted to ask if you have… you think there’s specific questions I should ask, because,
34 00:07:32.760 ⇒ 00:07:35.629 Amber Lin: Well, you guys have much more expertise in this.
35 00:07:35.830 ⇒ 00:07:39.590 Zoran Selinger: they are good. To me, really.
36 00:07:39.870 ⇒ 00:07:56.759 Zoran Selinger: like, this separation of… and those layers that you saw. So, like, AI for marketers, AI for customers, and then AI by customers. This is a really good way to look into… look into that.
37 00:07:57.050 ⇒ 00:08:10.059 Zoran Selinger: And I’m just… what I really want to talk… talk about with… with Robert is, how we serve to, like, each part of that, each part of that. So, I…
38 00:08:10.490 ⇒ 00:08:15.639 Zoran Selinger: definitely, Martech here is to,
39 00:08:16.240 ⇒ 00:08:19.270 Zoran Selinger: To collect good data, for sure.
40 00:08:19.440 ⇒ 00:08:31.599 Zoran Selinger: Because, you see, one of the… one of the biggest, biggest problems, and not one of them, but the biggest problem for AI adoption, proper AI adoption, is bad data.
41 00:08:32.630 ⇒ 00:08:37.889 Zoran Selinger: that companies collect. So, this is the main problem. We do solve that problem.
42 00:08:38.140 ⇒ 00:08:50.839 Zoran Selinger: just by doing our work. Yeah. We do solve that problem. So, I think, foundationally, this is… this is… this is good. Now, what kind of… because…
43 00:08:51.170 ⇒ 00:09:07.050 Zoran Selinger: I’m new here, I haven’t seen the AI part of our… of our implementations, so I haven’t seen this part. What do we, like, our AI enablement for our clients, how does that look like? I don’t know, I haven’t seen it.
44 00:09:07.940 ⇒ 00:09:18.890 Zoran Selinger: do we provide, like, internal chatbots for our clients to ask questions about the data itself?
45 00:09:19.600 ⇒ 00:09:22.920 Zoran Selinger: Yeah, I can give you a little bit of context. Okay, okay.
46 00:09:22.920 ⇒ 00:09:29.260 Amber Lin: for… on the ABC side, we made an internal chatbot for their customer service folks.
47 00:09:29.260 ⇒ 00:09:29.930 Zoran Selinger: Oh, okay.
48 00:09:29.930 ⇒ 00:09:36.550 Amber Lin: This is a little bit less data heavy, though we did put, because they need to ask about
49 00:09:36.680 ⇒ 00:09:45.139 Amber Lin: Who does… which technician does this zip code? So we help them put that data in a database, and then let them…
50 00:09:45.340 ⇒ 00:09:53.739 Amber Lin: So we have a text-to-SQL tool, and then use that to query the database and return results, and I think
51 00:09:54.090 ⇒ 00:09:55.110 Amber Lin: Casey, also?
52 00:09:55.110 ⇒ 00:09:58.060 Zoran Selinger: Fairly narrow, specific use case, right?
53 00:09:58.060 ⇒ 00:10:11.029 Amber Lin: Yes, exactly, and then Casey also used, our text to SQL for Eden. That should be a bit wider use case, because they want to ask about, some sales data.
54 00:10:11.030 ⇒ 00:10:23.539 Amber Lin: And they can also ask about SLA and operations, so that’s another use case on Eden, which should be broader. But overall, I think that is still quite a new…
55 00:10:23.610 ⇒ 00:10:29.729 Amber Lin: ability that we’re trying to get more clients to buy in on.
56 00:10:29.730 ⇒ 00:10:32.090 Zoran Selinger: That has internal tools, so that, right?
57 00:10:32.090 ⇒ 00:10:33.240 Amber Lin: Yes, yes.
58 00:10:33.240 ⇒ 00:10:35.979 Zoran Selinger: That’s this layer of…
59 00:10:36.100 ⇒ 00:10:49.100 Zoran Selinger: of, AI for marketers, or sales, or customer service. Do we have any examples of us being, shipping anything that customers
60 00:10:49.270 ⇒ 00:10:50.190 Zoran Selinger: use.
61 00:10:50.790 ⇒ 00:10:57.430 Amber Lin: No. So, because I think, especially because we… We currently…
62 00:10:58.360 ⇒ 00:11:04.410 Amber Lin: mostly just directly built solutions for our customers. We don’t… haven’t built anything
63 00:11:04.540 ⇒ 00:11:08.650 Amber Lin: That’s a product, per se, that’s PTC.
64 00:11:08.820 ⇒ 00:11:17.450 Amber Lin: And I don’t know if they plan to build that, unless one of our clients asks us to.
65 00:11:17.950 ⇒ 00:11:32.320 Zoran Selinger: Yeah, yeah, I understand, I understand, that makes sense. Yeah, so that’s… yeah, and I don’t know if, like, this… then this third part about… about AI by… by customers.
66 00:11:32.710 ⇒ 00:11:35.710 Zoran Selinger: Really the only way to affect this.
67 00:11:35.970 ⇒ 00:11:40.329 Zoran Selinger: At least… from what I could understand, is by SEO.
68 00:11:42.490 ⇒ 00:11:49.379 Amber Lin: They do some… sometimes… our company also does, like, AI engine optimization.
69 00:11:49.380 ⇒ 00:11:50.120 Zoran Selinger: Sweet.
70 00:11:50.120 ⇒ 00:11:59.350 Amber Lin: also appear on a lot of AI results. I think especially Utam and the AI team has been focusing on that.
71 00:11:59.480 ⇒ 00:12:07.499 Amber Lin: where we… they’ve been optimizing our websites for, for AI search, but as you know, like, there are… the…
72 00:12:07.770 ⇒ 00:12:12.140 Amber Lin: Research firm also said There’s not just…
73 00:12:12.400 ⇒ 00:12:17.190 Amber Lin: optimizing the words in your articles for AI, there’s a lot much.
74 00:12:17.610 ⇒ 00:12:19.740 Zoran Selinger: Much more, more to that. Yeah, yeah, yeah.
75 00:12:19.740 ⇒ 00:12:24.419 Amber Lin: Yeah, so… Should… we should list that out, but…
76 00:12:25.090 ⇒ 00:12:32.380 Zoran Selinger: Yeah, and you think… so you’re saying that some of the techniques are… techniques are outside of what SEO would be?
77 00:12:33.500 ⇒ 00:12:34.840 Amber Lin: Yes, so…
78 00:12:34.840 ⇒ 00:12:35.480 Zoran Selinger: Huh?
79 00:12:35.880 ⇒ 00:12:41.439 Amber Lin: So, for example, how… very direct example how people optimize for
80 00:12:41.570 ⇒ 00:12:47.609 Amber Lin: AI engine, so that’s AEO. This is the same… essentially the same as SEO, but…
81 00:12:47.730 ⇒ 00:12:53.279 Amber Lin: It’s how people optimize for AI search.
82 00:12:53.880 ⇒ 00:12:58.450 Amber Lin: So, let me see if I can find… The quote.
83 00:12:58.450 ⇒ 00:13:04.699 Zoran Selinger: They had never looked into it before. I know this is happening, but they never really looked at technique.
84 00:13:04.700 ⇒ 00:13:05.030 Amber Lin: code.
85 00:13:05.030 ⇒ 00:13:06.649 Zoran Selinger: That, that people do.
86 00:13:06.960 ⇒ 00:13:17.899 Zoran Selinger: I know a little bit about SEO from before, but the optimization for AI… I know it’s happening, but I don’t know exactly what people are doing.
87 00:13:18.370 ⇒ 00:13:23.139 Zoran Selinger: Basically, like, structured snippets and things like that, structured text.
88 00:13:23.140 ⇒ 00:13:23.540 Amber Lin: Yay!
89 00:13:23.540 ⇒ 00:13:30.190 Zoran Selinger: Are useful, for sure, but there’s probably a bunch of other stuff that’s happening there.
90 00:13:30.190 ⇒ 00:13:32.350 Amber Lin: Yeah, let’s see if I can…
91 00:13:32.670 ⇒ 00:13:39.560 Amber Lin: find the quote in there, I think that would be… That would be helpful.
92 00:13:40.150 ⇒ 00:13:48.520 Zoran Selinger: Basically, I’m thinking about how do we… how do we address that part of our… our clients? How can we,
93 00:13:48.620 ⇒ 00:13:49.450 Zoran Selinger: Guys…
94 00:13:50.370 ⇒ 00:13:57.100 Zoran Selinger: that’s the report. The report, is kind of showing us that marketers will, and clients will come to us.
95 00:13:57.100 ⇒ 00:13:59.040 Amber Lin: So weird.
96 00:13:59.040 ⇒ 00:14:02.420 Zoran Selinger: And they will ask about those 3 levels, basically.
97 00:14:02.490 ⇒ 00:14:03.750 Amber Lin: And…
98 00:14:03.940 ⇒ 00:14:12.569 Zoran Selinger: I’m just trying to understand what our scope is. Do we… Ever, for example, touch
99 00:14:12.800 ⇒ 00:14:22.799 Zoran Selinger: this part of AI by customers, or we are kind of limited to AI for marketers, and then maybe a little bit of
100 00:14:23.260 ⇒ 00:14:25.209 Zoran Selinger: AI for customers.
101 00:14:25.630 ⇒ 00:14:26.310 Zoran Selinger: From that.
102 00:14:26.310 ⇒ 00:14:26.790 Amber Lin: I see.
103 00:14:26.790 ⇒ 00:14:27.309 Zoran Selinger: I’m really eating.
104 00:14:27.310 ⇒ 00:14:32.019 Amber Lin: A lot of our clients don’t want us directly influencing the
105 00:14:32.430 ⇒ 00:14:38.940 Amber Lin: Like, the content, or the execution, or we’ve been limited in what we…
106 00:14:39.500 ⇒ 00:14:54.340 Amber Lin: do there, I believe. The only example I remember is on Insomnia, we were working directly with them with some of their email content, but we’ve always stayed on the data side, so.
107 00:14:54.920 ⇒ 00:15:01.300 Amber Lin: I… but I do think we can consult our clients on, say,
108 00:15:02.320 ⇒ 00:15:11.429 Amber Lin: what’s a… what’s a better way to do these? I just found this in… In the report, so…
109 00:15:11.650 ⇒ 00:15:15.709 Amber Lin: We do… a lot of people do this, and we do this pretty well.
110 00:15:16.830 ⇒ 00:15:19.199 Amber Lin: So there’s that, and then there’s also…
111 00:15:19.200 ⇒ 00:15:23.490 Zoran Selinger: Sorry, sorry, sorry, tell me this. Is this for us, for our website?
112 00:15:23.710 ⇒ 00:15:24.240 Zoran Selinger: Or for
113 00:15:25.240 ⇒ 00:15:30.010 Zoran Selinger: We don’t do this for clients, right? We do this for ourselves, we don’t do this for clients yet, but…
114 00:15:30.010 ⇒ 00:15:33.579 Amber Lin: But this is a way that we can help consult them, is, hey.
115 00:15:33.580 ⇒ 00:15:33.950 Zoran Selinger: Okay.
116 00:15:33.950 ⇒ 00:15:35.929 Amber Lin: You can do this better, of…
117 00:15:36.030 ⇒ 00:15:54.970 Amber Lin: we can do, like, a website audit for them, and then, like, these are more technical things that I don’t… like, most clients are not doing. I don’t know if we are doing that, but I think we’re certainly better posed to understand these things, because we do have an AI team.
118 00:15:55.120 ⇒ 00:16:01.019 Amber Lin: so that’s, like, this is the list of how you influence
119 00:16:02.010 ⇒ 00:16:09.820 Amber Lin: agents for clients. So, we can reference… we can reference this to say, hey, we… we’ve done this ourselves, or we can…
120 00:16:09.980 ⇒ 00:16:11.540 Amber Lin: Could help clients do that.
121 00:16:11.540 ⇒ 00:16:20.340 Zoran Selinger: Interesting, yeah. I wonder what… what… what Robert thinks, and you tell what… what they think about… Yeah. …that. Okay.
122 00:16:20.340 ⇒ 00:16:25.730 Amber Lin: Yeah. I really think Utah should be part of this conversation, because he does so much of our…
123 00:16:26.070 ⇒ 00:16:39.370 Amber Lin: of our AI team stuff, and this report is so surprisingly about AI. Yeah. Why don’t we propose that in our chat? I think it’s way too early for Robert, because this is the day after New Year’s Eve.
124 00:16:39.530 ⇒ 00:16:45.440 Zoran Selinger: Yeah, let’s just tell him he… we discussed it a little bit, maybe? He can watch the…
125 00:16:45.830 ⇒ 00:16:46.200 Zoran Selinger: Thank you.
126 00:16:46.200 ⇒ 00:16:47.600 Amber Lin: Utam should be…
127 00:16:47.740 ⇒ 00:16:52.820 Zoran Selinger: Yeah, and we rescheduled with Utam, sure. Israel.
128 00:16:53.150 ⇒ 00:17:03.710 Zoran Selinger: Cool, okay. I mean, I think that’s a good… I think I agree with you that, probably, since this is so AI-heavy, we should… we should have him in the conversation.
129 00:17:04.490 ⇒ 00:17:06.819 Amber Lin: Okay, I’ll send it in our little chat.
130 00:17:06.819 ⇒ 00:17:10.689 Zoran Selinger: Yeah, thanks for talking. Yeah, thank you, thank you. Have a good one.
131 00:17:10.690 ⇒ 00:17:11.960 Amber Lin: Talk soon. Bye-bye. Bye.