Meeting Title: AI Team Retro Date: 2025-09-05 Meeting participants: Samuel Roberts, Casie Aviles, Mustafa Raja


WEBVTT

1 00:01:28.490 00:01:30.400 Samuel Roberts: There we go. Can you guys hear me now?

2 00:01:31.240 00:01:32.060 Mustafa Raja: Yep.

3 00:01:32.590 00:01:33.770 Samuel Roberts: Alright, cool.

4 00:01:34.390 00:01:37.570 Samuel Roberts: Oh, my headphones are doing weird things, that’s why I’m messing up here.

5 00:01:38.440 00:01:42.019 Samuel Roberts: I was, like, talking and nothing was lighting up on my computer.

6 00:01:42.430 00:01:43.230 Mustafa Raja: Yeah.

7 00:01:45.070 00:01:48.909 Samuel Roberts: Yeah, sorry about that. I forgot that that’s a… it’s a different event, technically.

8 00:01:49.540 00:01:52.300 Samuel Roberts: than the others, so when I added the meeting.

9 00:01:52.650 00:01:55.669 Samuel Roberts: To the main 3 during the week, it didn’t add it to,

10 00:01:56.490 00:01:58.520 Samuel Roberts: The fourth one, or the last one on Friday.

11 00:02:00.350 00:02:05.480 Samuel Roberts: So I just went to go join it, and it wasn’t there, and I was like, oh no, hopefully they didn’t do the same thing, but… how are you guys doing?

12 00:02:06.160 00:02:07.460 Mustafa Raja: Yeah, I’m doing good.

13 00:02:07.980 00:02:08.610 Samuel Roberts: And…

14 00:02:08.610 00:02:10.130 Casie Aviles: Doing good, doing good.

15 00:02:10.139 00:02:11.029 Samuel Roberts: Cool, cool.

16 00:02:13.119 00:02:17.299 Samuel Roberts: Yeah, so how… let’s just chat quick about the week, and then we can maybe hop into…

17 00:02:18.159 00:02:19.659 Samuel Roberts: Linear if we need.

18 00:02:20.069 00:02:21.789 Samuel Roberts: How are you guys feeling about this week?

19 00:02:23.140 00:02:26.670 Mustafa Raja: Yeah, I guess… I guess it went good, I guess.

20 00:02:27.890 00:02:37.970 Mustafa Raja: Had, had, a bit of default work, and we were able to generate a report based on the signals that we have been trying to

21 00:02:38.180 00:02:40.499 Mustafa Raja: listened to.

22 00:02:40.500 00:02:41.590 Samuel Roberts: Oh, yeah, good.

23 00:02:41.590 00:02:47.360 Mustafa Raja: Yeah, I guess the, we still need a review from default about that.

24 00:02:47.780 00:02:50.110 Samuel Roberts: Yeah, I saw they moved.

25 00:02:50.340 00:02:51.080 Samuel Roberts: Alright.

26 00:02:52.320 00:02:53.110 Samuel Roberts: Good.

27 00:02:54.770 00:02:57.510 Mustafa Raja: And yeah, Interlude started again, so…

28 00:02:57.860 00:02:58.470 Samuel Roberts: Yep.

29 00:02:59.380 00:03:00.949 Mustafa Raja: That’ll be good, I guess.

30 00:03:02.270 00:03:05.600 Samuel Roberts: Definitely, yeah, yeah. I’m excited to keep working on that one.

31 00:03:07.700 00:03:09.160 Samuel Roberts: Casey, how about you?

32 00:03:09.760 00:03:17.190 Casie Aviles: Yeah, I guess for me, like, there’s just a lot… a lot of stuff that happened, like, outside of work that…

33 00:03:17.450 00:03:21.369 Casie Aviles: They’re kind of out of my control, so that kind of impacted

34 00:03:23.540 00:03:30.920 Casie Aviles: you know, like, the speed of me getting stuff out, but I think I’m… I think I’ve managed to get on track, at least.

35 00:03:31.130 00:03:33.490 Casie Aviles: Yeah, I wouldn’t say so.

36 00:03:34.420 00:03:41.250 Casie Aviles: Yeah, yeah, that’s… at least that’s good. Yeah, it’s also nice to see that.

37 00:03:41.930 00:03:46.330 Casie Aviles: One of the clients I’m working on, Insomnia Cookies, is also renewing and also.

38 00:03:46.330 00:03:46.820 Samuel Roberts: Yes.

39 00:03:46.820 00:03:47.550 Casie Aviles: to hear it.

40 00:03:47.990 00:03:52.730 Casie Aviles: that interlude, like, Mustafa’s client is also renewing, so that’s good.

41 00:03:53.140 00:03:53.530 Casie Aviles: Yeah.

42 00:03:54.830 00:03:56.369 Casie Aviles: What else? Yeah.

43 00:03:57.350 00:04:02.279 Casie Aviles: Yeah, it’s nice that we have… we were able to do at least some…

44 00:04:02.470 00:04:09.819 Casie Aviles: internal stuff, I believe, this week. Like, I know there’s… it’s not a lot still, but I think it’s nice that we were able to push some.

45 00:04:10.200 00:04:12.660 Casie Aviles: And, like, last week, last few weeks, we’re…

46 00:04:12.660 00:04:13.990 Samuel Roberts: Yeah, exactly.

47 00:04:14.150 00:04:20.940 Samuel Roberts: And I’ve got the… I didn’t get as much done yesterday as I was hoping to on the monorepo, but I got some stuff.

48 00:04:21.060 00:04:33.320 Samuel Roberts: Where it’s… it’s… it’s probably live on a link on Heroku, and so I can probably put something together. I need to update, like, README and maybe some cursor rules and stuff so we can all jump into it.

49 00:04:33.470 00:04:36.600 Samuel Roberts: But, yeah.

50 00:04:37.530 00:04:40.079 Casie Aviles: Nice. I think, I think we’re making progress there, too.

51 00:04:40.750 00:04:47.060 Samuel Roberts: Cool. And then ABC, it also looked like we… you got the, the database completed?

52 00:04:47.650 00:04:48.150 Samuel Roberts: And I saw.

53 00:04:48.150 00:04:49.270 Casie Aviles: Yes.

54 00:04:49.270 00:04:51.269 Samuel Roberts: And the testing looked good.

55 00:04:51.790 00:04:53.019 Samuel Roberts: Another quick test.

56 00:04:54.590 00:05:00.769 Casie Aviles: Yeah, those are just quick tests, but I don’t know, I’m not expecting it to, like, be 100%, like.

57 00:05:01.290 00:05:08.050 Casie Aviles: But clean, because even, like, the source data was… I found some inconsistencies, like.

58 00:05:08.940 00:05:10.570 Samuel Roberts: Yeah.

59 00:05:10.570 00:05:11.330 Casie Aviles: Yes.

60 00:05:12.150 00:05:17.780 Samuel Roberts: Yeah, I don’t really know how… I wasn’t obviously here for, like, the start of ABC…

61 00:05:18.330 00:05:28.889 Samuel Roberts: So, like, I don’t know exactly how it went, you know, taking their data, putting it into, like, the sheets, and doing the master sheet and everything, but it seems like their data is just scattered a lot, to begin.

62 00:05:28.890 00:05:29.620 Casie Aviles: Yeah.

63 00:05:30.010 00:05:32.850 Samuel Roberts: Which is… you know.

64 00:05:33.040 00:05:40.259 Samuel Roberts: Makes it difficult for us, makes it difficult for them and everything, so… hopefully this database, at least moving forward, will,

65 00:05:40.800 00:05:45.980 Samuel Roberts: Will be a good, like, source of truth we can keep updated, maybe? And get them to use, and…

66 00:05:46.980 00:05:47.570 Casie Aviles: Yes.

67 00:05:48.370 00:05:54.180 Samuel Roberts: Yeah, we had the call, with ABC yesterday.

68 00:05:54.770 00:05:59.369 Samuel Roberts: And I gave them, like, a little rundown of the proposal for, like, the new UI.

69 00:06:00.400 00:06:01.959 Casie Aviles: Hmm. Nice.

70 00:06:02.150 00:06:03.979 Casie Aviles: And so, like, the…

71 00:06:04.110 00:06:09.369 Samuel Roberts: the two, people on the call, but there were a couple… there was one guy who wasn’t there that I think is the, like.

72 00:06:09.510 00:06:18.360 Samuel Roberts: decision maker. So, Utam’s gonna have to meet with him and, like, talk pricing and stuff, but I’m hoping they’re… they’re excited to go for it, because I think…

73 00:06:18.530 00:06:23.410 Samuel Roberts: There’s a lot we can add, from, like, the transcript side once we get those in.

74 00:06:23.550 00:06:26.200 Samuel Roberts: That will be better…

75 00:06:26.960 00:06:31.449 Samuel Roberts: to visualize on a different UI than just with Google Chat at this point, you know?

76 00:06:31.450 00:06:32.010 Casie Aviles: That’s true.

77 00:06:32.650 00:06:42.919 Samuel Roberts: So I’m hoping… I’m hoping they want to go for it, because then I think… I don’t know if you saw the message I was going back and forth with you, Tom, but, like, I definitely want to lean into Copilot Kit and pick one of the frameworks to run with.

78 00:06:43.460 00:06:46.880 Samuel Roberts: So that we can, like, move, because I think…

79 00:06:47.020 00:06:50.600 Samuel Roberts: The NDDI workflow is massive, you know what I mean?

80 00:06:50.880 00:06:51.760 Casie Aviles: It is, yeah.

81 00:06:51.760 00:07:03.210 Samuel Roberts: And so, like, it just becomes, like, how to, you know, making changes to that becomes harder, and, like, I think if we can maybe… and this might be a little bit more if they want to go for it, I don’t want to just re-architect for the sake of re-architecting.

82 00:07:03.210 00:07:04.740 Casie Aviles: But.

83 00:07:04.740 00:07:13.799 Samuel Roberts: you know, pulling out some of that, like, higher level functionality, maybe some of, like, the more routing stuff, so that it can, like… CopilotK can handle it, and then it hits an NAD endpoint that does…

84 00:07:13.960 00:07:20.670 Samuel Roberts: a little bit of lookup, or whatever it is, for Andy, so that it might be a little bit more manageable.

85 00:07:21.160 00:07:26.640 Samuel Roberts: I started digging a little bit yesterday into some of those other frameworks, just to get a sense of them.

86 00:07:26.840 00:07:29.399 Samuel Roberts: And kind of their maturity level, and…

87 00:07:29.640 00:07:32.170 Samuel Roberts: How easy they are to implement.

88 00:07:32.290 00:07:39.780 Samuel Roberts: The biggest thing… is… the difference between Python versus TypeScript.

89 00:07:40.500 00:07:41.180 Mustafa Raja: Okay.

90 00:07:41.180 00:07:48.180 Samuel Roberts: that kind of dictates… my leaning is always TypeScript, because I know that much better than Python.

91 00:07:48.840 00:07:56.150 Samuel Roberts: But I don’t want to make that necessarily the determining factor. However, if we’re building, like, a Next app, for example.

92 00:07:56.690 00:08:00.050 Samuel Roberts: it might be easier to do everything in TypeScript, but…

93 00:08:00.050 00:08:00.660 Mustafa Raja: Hmm.

94 00:08:01.090 00:08:12.499 Samuel Roberts: I don’t know, I wanted to hear your thoughts a little bit, what your experiences are with, you know, Python versus TypeScript in general, how you guys… and I know, Mustafa, you’ve had some experience with some of these frameworks.

95 00:08:12.500 00:08:21.480 Mustafa Raja: Yeah, yeah, yeah. So, so far, I mean, it’s been… it’s been quite, quite some time since I’ve worked on those frameworks, but…

96 00:08:21.480 00:08:37.150 Mustafa Raja: Last time I worked on them, there’s always some… some libraries or some modules missing in TypeScript, or we can say TypeScript is always some… some ways behind the Python version of the library.

97 00:08:37.370 00:08:39.049 Samuel Roberts: This is, like, Langchain, right? Specifically?

98 00:08:39.059 00:08:40.599 Mustafa Raja: Yeah, Lang Chen and Laman.

99 00:08:40.600 00:08:49.699 Samuel Roberts: Yeah, Llama Index 2, okay, yeah. Because I was looking at… there’s another one called Mastra, or Mastra, I don’t know, that is just TypeScript, so it’s not trying.

100 00:08:50.530 00:08:51.290 Mustafa Raja: Yeah.

101 00:08:51.290 00:08:53.520 Samuel Roberts: And so, I was watching a couple videos, because it’s a.

102 00:08:53.520 00:08:54.510 Mustafa Raja: I mean, it’s…

103 00:08:54.510 00:08:56.050 Samuel Roberts: Different, but go ahead.

104 00:08:56.470 00:09:07.110 Mustafa Raja: Yeah, yeah, if it’s, if it’s compatible, master is compatible with, what’s it called? Copilot, then, I guess, we should do that, since it’s based on only TypeScript.

105 00:09:07.530 00:09:19.220 Samuel Roberts: That’s kind of my… my thought, but again, like, I am a TypeScript guy more than a Python guy, so I don’t want to, like, just make that decision without looking at them a little bit closer. So I watched a couple of

106 00:09:19.550 00:09:26.920 Samuel Roberts: of, like, examples and stuff, but it seems like it’s pretty good. And I think because, you know, if we were just building back-end things.

107 00:09:28.420 00:09:33.320 Samuel Roberts: you know, something that was going to connect to Google Chat, just like Andy does, maybe Python would make sense for that.

108 00:09:33.510 00:09:49.799 Samuel Roberts: But if we’re gonna have to build a UI, and have a next app, or, you know, however we want it, I think it might make sense to have one language, but, I mean, just regardless of the frameworks, how do you guys feel, like, JavaScript, TypeScript versus Python from your abilities, you know?

109 00:09:51.150 00:09:52.850 Samuel Roberts: What do you… what do you do? What do you feel more comfortable?

110 00:09:52.850 00:09:53.800 Mustafa Raja: I’m,

111 00:09:53.910 00:10:07.959 Mustafa Raja: Yeah, I’m pretty comfortable with both of those. I feel, I feel, if we… if we do Python, we are kind of ripping off the, benefit that we would get from the monorepo itself.

112 00:10:08.350 00:10:13.879 Mustafa Raja: Sure. It’s going to be the same thing that we have currently.

113 00:10:14.310 00:10:34.350 Mustafa Raja: So, yeah. That’s a good point, yeah. I mean, we did monorepo just so we can, firstly, auto-deploy, have back-end the same thing, so we do not have to deploy a backend once, and then test the front end, and then deploy that.

114 00:10:34.350 00:10:37.099 Mustafa Raja: No, that’s a good point. So it defeats that purpose.

115 00:10:38.170 00:10:39.579 Mustafa Raja: Other than that…

116 00:10:40.310 00:10:54.050 Mustafa Raja: I don’t know, it’s just that, in my experience, Python is always, some steps ahead of TypeScript in terms of modules. That’s the only type breaker.

117 00:10:54.680 00:11:00.369 Samuel Roberts: Yeah, I… I hear that too, because when I was doing lane training, it was a very similar experience.

118 00:11:00.550 00:11:14.299 Samuel Roberts: But that was, you know, maybe a year ago, so, yeah, so I definitely want to look into Langchain a little bit more, too, just because I think that’s probably advanced, but I’m definitely going to spend some time with Monstra and see, because I saw a lot of,

119 00:11:14.490 00:11:19.559 Samuel Roberts: videos with CopilotKit and Mastra and, like, the teams together, so they seem to be working.

120 00:11:19.750 00:11:21.090 Samuel Roberts: Pretty closely.

121 00:11:21.090 00:11:32.489 Mustafa Raja: Yeah, I guess for them, it’s good too, because, CopilotKit is obviously in TypeScript, so they would also want something in TypeScript that’s more manageable.

122 00:11:32.830 00:11:34.080 Samuel Roberts: That’s true, that’s true.

123 00:11:34.830 00:11:38.550 Samuel Roberts: Yeah, Casey, any thoughts on Python versus TypeScript from your angle?

124 00:11:39.940 00:11:41.919 Casie Aviles: I mean, for me, I’m…

125 00:11:42.220 00:11:49.439 Casie Aviles: the opposite, like, I’m more of a Python… yeah, I learned Python more than I learned TypeScript.

126 00:11:49.770 00:11:50.440 Samuel Roberts: Okay.

127 00:11:50.830 00:12:00.540 Casie Aviles: I think it’s still, you know, I don’t mind, like, it’s fine if we think that TypeScript, or, like, Mastra and CopilotKit works.

128 00:12:00.820 00:12:05.020 Casie Aviles: Much better, and if we can… I think if we can…

129 00:12:05.310 00:12:10.670 Casie Aviles: Create better products that way, or, like, we can do stuff much…

130 00:12:10.890 00:12:13.559 Casie Aviles: Better that way, then it’s fine also for me.

131 00:12:13.940 00:12:14.690 Samuel Roberts: Okay, yeah.

132 00:12:15.060 00:12:15.840 Mustafa Raja: Hmm.

133 00:12:16.190 00:12:22.849 Samuel Roberts: Yeah, I mean, it’s a good opportunity to get, you know, more experience with TypeScript as well, because, like, if there’s any amount of front-end work, or any amount of…

134 00:12:23.020 00:12:26.380 Samuel Roberts: You know, platform work that we’re gonna have, it’s probably gonna be that.

135 00:12:26.760 00:12:28.929 Samuel Roberts: Which is, like, the reason I’m…

136 00:12:29.070 00:12:42.190 Samuel Roberts: a kind of full-stack TypeScript guy in my mind, because I was always doing projects kind of solo or with a designer, and so, like, it didn’t make sense to be like, oh, yeah, let’s build our backend in something different, and then I have to jump

137 00:12:42.980 00:12:45.710 Samuel Roberts: Contact switch and everything, so,

138 00:12:45.900 00:12:50.210 Samuel Roberts: Yeah, okay, good. I just, yeah, I wanted to get a little sense of that from you guys,

139 00:12:50.370 00:12:58.839 Samuel Roberts: But I think, you know, I… I love TypeScript, because I can work in the web, and I’m a big… I like web apps, and building things, and, you know…

140 00:12:59.180 00:13:08.090 Samuel Roberts: I have just migrated to that overall, over time, but again, I don’t want to just decide that, so I’m going to do a little more research, probably put together.

141 00:13:08.090 00:13:09.100 Mustafa Raja: Yeah. Spike.

142 00:13:09.170 00:13:13.250 Samuel Roberts: But I try to understand, because Monster looks like it’s a slightly different

143 00:13:14.630 00:13:23.300 Samuel Roberts: like, design philosophy than Langchain. But it’s, like, similar in some ways, but they use different terminology, and I’m just trying to understand how it all goes together, but…

144 00:13:23.520 00:13:27.879 Samuel Roberts: At least for this project, maybe it’s worth trying it out. So.

145 00:13:29.010 00:13:33.619 Samuel Roberts: Because, yeah, sometimes just doing something and picking something is better than, you know.

146 00:13:33.880 00:13:36.029 Samuel Roberts: Picking the right or wrong thing, even.

147 00:13:37.480 00:13:39.500 Samuel Roberts: Cool, alright. Yeah, go ahead.

148 00:13:39.500 00:13:40.760 Casie Aviles: It turns out that

149 00:13:40.910 00:13:48.730 Casie Aviles: it doesn’t… it’s not that good, and I guess we could just look for a different one. At least we tried something.

150 00:13:48.730 00:14:00.989 Samuel Roberts: Exactly, exactly, you know what I mean? So I don’t want to spend too much time on it, but there were a couple videos I was watching yesterday that I want to finish up today, and maybe do a demo or something with it, because when I did the Copilot Kit demos, I was either just hitting N8N,

151 00:14:01.290 00:14:06.540 Samuel Roberts: Or I was just using OpenAI just straight to the LLM, which is…

152 00:14:06.810 00:14:11.300 Samuel Roberts: Workable. But obviously, if you’re trying to do more complex things.

153 00:14:11.540 00:14:13.279 Samuel Roberts: The frameworks help out with that.

154 00:14:13.750 00:14:19.610 Samuel Roberts: And then there’s stuff about, observability and tracing that I was having a really hard time getting set up.

155 00:14:19.610 00:14:20.010 Casie Aviles: neither.

156 00:14:20.010 00:14:34.310 Samuel Roberts: pilot kit by itself, and obviously NADN has its own issues with that, but so I gotta do a little bit more research into how Maestra and Langchain handle that to make sure that it’s something good that we can pull the data out of for, you know, each trace and do a little bit better.

157 00:14:35.030 00:14:37.040 Samuel Roberts: Better job with that.

158 00:14:38.120 00:14:45.239 Samuel Roberts: So that’s kind of where I’m at with it in general, but I definitely wanted to hear your thoughts on those, but I’ll try to get that spike together.

159 00:14:45.460 00:14:48.100 Samuel Roberts: After I… after I dig in a little bit to Mastra, so…

160 00:14:50.230 00:14:51.400 Casie Aviles: Okay. Good.

161 00:14:51.400 00:14:54.209 Samuel Roberts: Good. So what does today look like for you guys, then?

162 00:14:56.140 00:14:56.930 Samuel Roberts: Sweet.

163 00:14:56.930 00:15:07.869 Mustafa Raja: Yeah, for me, some default work, some ABC work, and then for Interlude, maybe we can work on the FigJam that we started.

164 00:15:08.440 00:15:21.789 Samuel Roberts: Yeah, I may or may not have time today for that, but feel free to jump in without me as well, that’s kind of the point, you know? So if you want to, like, diagram a couple of the other ones, like, just ping me and let me know you’re doing it, and I’ll try to take a look.

165 00:15:22.410 00:15:27.079 Mustafa Raja: Okay. Yeah, send me the link in the, what’s it called, Slack, please?

166 00:15:27.290 00:15:33.179 Samuel Roberts: Oh, totally, yeah, sorry, I forgot we were on the call when I did that. Yeah, I’ll do that, I’ll do that right after this if I…

167 00:15:33.660 00:15:37.310 Samuel Roberts: Can’t find where I have it open, because I have so many tabs open sometimes.

168 00:15:37.310 00:15:37.940 Mustafa Raja: Yeah.

169 00:15:38.180 00:15:42.100 Samuel Roberts: I’ll grab it, yeah, I’ll definitely send it to you. Okay. And then, Casey, how about you?

170 00:15:43.400 00:15:49.449 Casie Aviles: For me, it’s just… I’m just going to wrap up on Insomnia Cookies, like, the remaining…

171 00:15:49.580 00:15:50.989 Casie Aviles: Automation that they have.

172 00:15:51.670 00:15:59.510 Casie Aviles: And then for ABC… Yeah, just triage tickets that I haven’t gotten around to yet.

173 00:16:00.680 00:16:01.090 Samuel Roberts: Okay.

174 00:16:01.090 00:16:04.100 Casie Aviles: Yeah, I think that’s… that’s it. There’s just some…

175 00:16:04.410 00:16:08.170 Casie Aviles: Stuff coming in from the AI team, internal work.

176 00:16:08.620 00:16:09.790 Samuel Roberts: Saw some of that, yeah.

177 00:16:10.240 00:16:13.650 Casie Aviles: So I guess, yeah, I’ll try to work on that.

178 00:16:14.150 00:16:14.940 Casie Aviles: as well.

179 00:16:15.480 00:16:23.740 Samuel Roberts: Yeah, I would say, doesn’t have access to my email, I can probably start pulling them all from sent email. Okay, cool, I was just catching up on what your Tom just said.

180 00:16:24.130 00:16:27.279 Samuel Roberts: Yeah, I would say…

181 00:16:30.810 00:16:32.910 Samuel Roberts: Yeah, okay, I at least, like…

182 00:16:33.520 00:16:35.829 Samuel Roberts: It might be a good thing now that… because, like.

183 00:16:36.340 00:16:41.600 Samuel Roberts: It’s good to have the ticket there, and it’s good to have the context, but maybe just, like, spend a minute

184 00:16:41.740 00:16:44.360 Samuel Roberts: Specking out, like, what the, kind of.

185 00:16:44.820 00:16:47.040 Samuel Roberts: Work is gonna be for that.

186 00:16:47.800 00:16:49.190 Casie Aviles: Hmm, if that makes sense?

187 00:16:49.260 00:16:57.179 Samuel Roberts: That way, like, it’s… because I want to get better about this, and this is a whole other broader thing about, like, having the tickets groomed a little bit better.

188 00:16:57.320 00:16:58.360 Samuel Roberts: So that…

189 00:16:58.360 00:16:59.010 Casie Aviles: Yeah.

190 00:16:59.010 00:17:10.230 Samuel Roberts: there’s a little more context for everyone looking at it, not just the person doing it, obviously, but they definitely need it too. But that way, you know, if someone needs to jump in, or someone can help with a subtask or something.

191 00:17:10.730 00:17:14.129 Samuel Roberts: So, I would say, yeah, so what did Hustle said?

192 00:17:14.130 00:17:16.109 Casie Aviles: Snippets in HubSpot.

193 00:17:16.319 00:17:28.979 Samuel Roberts: I would just say, yeah, just maybe, like, you know, a quick brain dump of, like, what you think is needed to do it, and just, like, get that into linear as, like, part of a description or something, and then that way… because, like.

194 00:17:29.060 00:17:36.600 Samuel Roberts: The other side of it is we… we can’t always just, like, jump on whatever someone asks us to do when there’s other client work to do.

195 00:17:37.350 00:17:42.180 Samuel Roberts: So, I want to make sure that, like, we’re still coordinating tickets pretty well.

196 00:17:42.380 00:17:55.540 Samuel Roberts: And definitely having a better sense of, like, okay, how complex is this ticket? You know, is it something we can just, like, quickly get out? Are we waiting on other things? You know, that whole thing, I think, would be a little helpful.

197 00:17:57.030 00:17:58.480 Casie Aviles: Yeah, yeah, definitely.

198 00:17:58.480 00:17:59.000 Samuel Roberts: So…

199 00:17:59.000 00:18:02.880 Casie Aviles: And then also, yeah. Like, a quick to-do list there, I guess, for…

200 00:18:02.880 00:18:04.430 Samuel Roberts: Exactly, exactly, yeah.

201 00:18:04.630 00:18:09.210 Samuel Roberts: Yeah, that’s… exactly. And I think that’s probably a good way to go for these,

202 00:18:09.580 00:18:14.429 Samuel Roberts: kind of, like, little requests, that we get, that way we’re not, like.

203 00:18:14.670 00:18:19.609 Samuel Roberts: We’re not ignoring them, obviously, but we’re also not just, like, dropping everything and doing them.

204 00:18:19.940 00:18:26.770 Samuel Roberts: We gotta find a good balance of… that’s kind of our struggle as the AI team, is the balance of, like, client work and internal work.

205 00:18:26.770 00:18:27.200 Mustafa Raja: Yeah.

206 00:18:27.200 00:18:29.629 Samuel Roberts: And so I want to make sure we’re…

207 00:18:29.920 00:18:33.709 Samuel Roberts: Prioritizing things well without letting too much fall through the cracks.

208 00:18:34.350 00:18:35.440 Samuel Roberts: But, we’re getting there.

209 00:18:35.440 00:18:36.000 Mustafa Raja: Yep.

210 00:18:37.280 00:18:37.970 Samuel Roberts: Right.

211 00:18:39.510 00:18:40.260 Samuel Roberts: Cool.

212 00:18:41.040 00:18:46.989 Samuel Roberts: Trying to think what else, I have a little bit of a lighter day today, I don’t think I’m gonna be…

213 00:18:47.290 00:18:54.120 Samuel Roberts: on, a ton, I’m not sure yet, but, ping me if you need me. I’ll probably be…

214 00:18:54.840 00:18:57.100 Samuel Roberts: At the meeting at 11.

215 00:18:57.250 00:19:05.590 Samuel Roberts: My time is, like, in almost an hour. Okay, yeah, so let me know if I can help out with anything. Otherwise, I’ll see you guys on Slack, yeah.

216 00:19:07.230 00:19:08.730 Casie Aviles: Alright, thank you guys.

217 00:19:08.730 00:19:09.050 Samuel Roberts: Alright.

218 00:19:09.050 00:19:10.140 Mustafa Raja: Thank you. Yep.

219 00:19:10.280 00:19:13.370 Samuel Roberts: Have a good one, and if I don’t talk to you guys, have a good weekend.

220 00:19:14.320 00:19:16.370 Mustafa Raja: Yeah, alright. You too, bye.

221 00:19:16.410 00:19:17.340 Samuel Roberts: Thanks, bye.

222 00:19:17.340 00:19:17.950 Mustafa Raja: Bye.