Meeting Title: ABC Project Working Session Date: 2026-01-14 Meeting participants: Casie Aviles, Amber Lin, Mustafa Raja, Sam Roberts
WEBVTT
1 00:00:54.520 ⇒ 00:00:55.750 Amber Lin: Hello!
2 00:01:02.600 ⇒ 00:01:08.840 Amber Lin: Okay. I think while the others are not here yet, I can show you what I got from the dispatch meeting.
3 00:01:09.050 ⇒ 00:01:14.920 Amber Lin: I think we’re almost good to just add it in. Let me show you.
4 00:01:14.920 ⇒ 00:01:17.450 Mustafa Raja: I was able… oh, not this one.
5 00:01:17.560 ⇒ 00:01:23.070 Amber Lin: This… this one. So I consolidated it.
6 00:01:23.500 ⇒ 00:01:33.139 Amber Lin: And so, essentially, what I… oh, they… they… and they were able to fill in the gas that was missing, so I think we have this
7 00:01:33.280 ⇒ 00:01:36.599 Amber Lin: I got all the different…
8 00:01:37.400 ⇒ 00:01:47.120 Amber Lin: columns from them. So for example, trades… I think their tradesman only does residential, so I just added another column since we have that.
9 00:01:47.700 ⇒ 00:01:48.500 Casie Aviles: Yeah.
10 00:01:50.500 ⇒ 00:01:55.210 Amber Lin: I think we might need to name the columns a little differently based on how they ask
11 00:01:55.440 ⇒ 00:02:03.600 Amber Lin: the questions, but… Do you think this many… columns is okay. Like, there’s…
12 00:02:03.600 ⇒ 00:02:06.940 Casie Aviles: So that’ll be until the home zip column, right?
13 00:02:07.420 ⇒ 00:02:08.989 Amber Lin: Yeah, until the homes up column.
14 00:02:10.389 ⇒ 00:02:12.829 Casie Aviles: Okay, yeah, I think we can work with this.
15 00:02:12.830 ⇒ 00:02:21.559 Amber Lin: Okay, so, and that’s… I believe that’s all the mechanical texts, essentially all the dispatch… dispatch techs.
16 00:02:21.560 ⇒ 00:02:23.199 Casie Aviles: Hmm, I see.
17 00:02:23.920 ⇒ 00:02:28.839 Amber Lin: I’m gonna remove the… okay, that’s better.
18 00:02:30.060 ⇒ 00:02:32.580 Amber Lin: Okay, and then…
19 00:02:36.870 ⇒ 00:02:42.049 Sam Roberts: Sorry, I was… Hello! My Zoom, I had to reboot my computer and things weren’t logged in right.
20 00:02:42.510 ⇒ 00:02:49.410 Amber Lin: Yeah, I had that this morning, too. It’s, the more data I download, the slower it goes, so…
21 00:02:49.410 ⇒ 00:02:51.820 Sam Roberts: Yeah. Yep, I hear that.
22 00:02:52.310 ⇒ 00:03:01.559 Amber Lin: Cool. I’m just talking with Casey about the mechanical texts. I think we have pretty much… this is good. I think I’ll…
23 00:03:01.760 ⇒ 00:03:03.240 Sam Roberts: Also, just…
24 00:03:03.740 ⇒ 00:03:15.220 Amber Lin: Casey, I’ll send you this… This one as well. So these are the questions they ask in app.
25 00:03:15.220 ⇒ 00:03:16.160 Casie Aviles: I’ve read it.
26 00:03:16.160 ⇒ 00:03:30.910 Amber Lin: Like, they ask it like this, so can I book a service call on this person? So, I don’t know how the text to SQL will translate, but for example, when they ask service call, what they’re actually asking is,
27 00:03:31.870 ⇒ 00:03:40.199 Amber Lin: they’re asking about, I want to book a service call on this
28 00:03:40.400 ⇒ 00:03:45.950 Amber Lin: this person, who is a HVAC, I think on the HVAC.
29 00:03:46.380 ⇒ 00:03:54.160 Amber Lin: person. So, I might ask them if they can… more, more here, but…
30 00:03:55.150 ⇒ 00:04:03.400 Amber Lin: like, we might need to word these differently so that the text-to-SQL actually works, but I’ll put that in a ticket for you. Do you think this is doable?
31 00:04:05.430 ⇒ 00:04:10.300 Casie Aviles: Yeah, I’m just… I think… I think it should be doable,
32 00:04:10.440 ⇒ 00:04:16.430 Casie Aviles: Just, you know, we could just, basically update the instructions.
33 00:04:16.630 ⇒ 00:04:21.019 Casie Aviles: For… so that when… when we’re generating the SQL,
34 00:04:21.180 ⇒ 00:04:22.600 Casie Aviles: They should take a look at…
35 00:04:23.180 ⇒ 00:04:27.560 Casie Aviles: For now, if it’s the primary column, then we can just, you know, tell the…
36 00:04:29.190 ⇒ 00:04:32.110 Casie Aviles: Yeah, you can generate a SQL that will look at this
37 00:04:32.420 ⇒ 00:04:38.049 Casie Aviles: Primary column and take a look at service, and then, you know, it’s just a bunch of conditions that we’ll need to do.
38 00:04:38.470 ⇒ 00:04:41.290 Casie Aviles: I think it should be doable, yeah.
39 00:04:41.290 ⇒ 00:04:44.840 Amber Lin: Cool, okay, sounds good. Let me…
40 00:04:45.650 ⇒ 00:04:49.169 Sam Roberts: like, yvette just emailed back, I don’t know if you guys saw.
41 00:04:49.430 ⇒ 00:04:51.100 Sam Roberts: Like, literally 4 minutes ago.
42 00:04:51.230 ⇒ 00:04:53.960 Sam Roberts: They were out of town Monday and Tuesday, both Ernenese.
43 00:04:55.360 ⇒ 00:04:56.000 Casie Aviles: So.
44 00:04:56.000 ⇒ 00:05:03.119 Sam Roberts: they’re catching up. He said they’ll review everything, expect to respond by the end of the week. If we decide it’s gonna be helpful, we’ll let you know, so hopefully they’ll get a chance to play with it.
45 00:05:06.030 ⇒ 00:05:10.200 Casie Aviles: Okay, yeah, okay, nice. Yeah, I did chat with Janice as well.
46 00:05:10.800 ⇒ 00:05:11.380 Sam Roberts: Okay.
47 00:05:11.870 ⇒ 00:05:17.269 Casie Aviles: Briefly, I just told her, like, if she needs a call, then, yeah, she said the same thing, so…
48 00:05:17.630 ⇒ 00:05:21.250 Sam Roberts: Cool, okay, that’s good. At least there’s some… some context now. Okay.
49 00:05:23.360 ⇒ 00:05:24.900 Casie Aviles: Great.
50 00:05:25.070 ⇒ 00:05:35.510 Sam Roberts: Is there… I mean, I guess I’m happy to hang out and work. I don’t know, I have some other stuff I’m going… kind of in the middle of with Lilo, but, is there something I can help with right now, or I can just hang on if you need me.
51 00:05:35.890 ⇒ 00:05:41.499 Amber Lin: Yeah, I think Wilson, while you’re here, let’s just take a quick look at the Gantt chart.
52 00:05:41.830 ⇒ 00:05:50.609 Amber Lin: Make sure we’re on track, and then if there’s nothing, like, high level that we need you to set directions on, I think we can stay on and keep working.
53 00:05:50.950 ⇒ 00:05:51.880 Sam Roberts: Cool, okay.
54 00:05:51.880 ⇒ 00:05:57.900 Amber Lin: Yeah, let me share, since I have it here already. So, we’re at.
55 00:05:58.220 ⇒ 00:06:01.600 Sam Roberts: We’re on the 14th, midweek. Yes.
56 00:06:01.600 ⇒ 00:06:04.560 Amber Lin: Let’s see… so we have…
57 00:06:05.760 ⇒ 00:06:10.729 Amber Lin: what’s in progress, or whatever is before, so… Yeah.
58 00:06:10.730 ⇒ 00:06:14.779 Sam Roberts: Programmatic access is now done, right? Or we have access, at least.
59 00:06:15.100 ⇒ 00:06:20.130 Mustafa Raja: Yeah, yeah, yeah, that’s done. Where is that? This one, 15, 15.
60 00:06:20.130 ⇒ 00:06:20.700 Sam Roberts: Yeah.
61 00:06:20.700 ⇒ 00:06:21.690 Amber Lin: Okay.
62 00:06:22.380 ⇒ 00:06:24.420 Amber Lin: Okay, so we have that, that’s great.
63 00:06:26.910 ⇒ 00:06:29.450 Mustafa Raja: And then let’s mark 27 as done.
64 00:06:30.210 ⇒ 00:06:31.500 Amber Lin: 27?
65 00:06:32.350 ⇒ 00:06:33.140 Sam Roberts: Yeah, cool.
66 00:06:33.460 ⇒ 00:06:37.450 Amber Lin: Cool. Should we… have we done a review of that? I know you sent a…
67 00:06:37.590 ⇒ 00:06:48.169 Amber Lin: I know you sent something there, so I’ll go review that, but it’s migrated over, so I’ll say that’s done. And technically, this is in progress… I’ll put this as, like, started.
68 00:06:48.670 ⇒ 00:06:49.130 Sam Roberts: Yeah.
69 00:06:49.130 ⇒ 00:06:52.490 Mustafa Raja: Yeah, and for cancellations.
70 00:06:52.950 ⇒ 00:06:55.170 Mustafa Raja: I believe,
71 00:06:58.920 ⇒ 00:07:04.990 Mustafa Raja: Oh, so, so, I’m trying to understand what’s there to do now, you know?
72 00:07:05.570 ⇒ 00:07:14.210 Amber Lin: Yeah, I wanted to talk about it later with you two, because I think there’s some stuff to be done.
73 00:07:14.490 ⇒ 00:07:18.160 Amber Lin: Is… do you need Sam to be there when we talk about.
74 00:07:18.160 ⇒ 00:07:22.690 Mustafa Raja: Yeah, yeah, that would be nice, if Sam… if Sam is also there.
75 00:07:22.870 ⇒ 00:07:24.449 Mustafa Raja: Yeah, that would be nice.
76 00:07:24.690 ⇒ 00:07:30.659 Amber Lin: Cool, okay, let’s go through this, and then we’ll go back to talk about cancellations. That’s top of mind for me, too.
77 00:07:31.040 ⇒ 00:07:31.890 Mustafa Raja: Cool.
78 00:07:32.140 ⇒ 00:07:36.379 Amber Lin: Yeah. How are we on this, eval dataset?
79 00:07:37.770 ⇒ 00:07:43.050 Mustafa Raja: Yeah, I have bring, I have brought in, what’s it called, stuff from, Snowflake.
80 00:07:43.210 ⇒ 00:07:49.109 Mustafa Raja: I’m just going to clean it a little bit and have team review it.
81 00:07:49.250 ⇒ 00:07:53.139 Mustafa Raja: That’s pretty much it. I’ll have it done by… by today.
82 00:07:53.720 ⇒ 00:08:00.090 Amber Lin: Okay. What about this one? Is this blocked by anything? Man’s right versus NF?
83 00:08:00.090 ⇒ 00:08:09.739 Mustafa Raja: I think, validation… I think it’s, we have already done this, if I’m wrong.
84 00:08:09.740 ⇒ 00:08:17.729 Sam Roberts: Yeah, I mean, I think we’re just… we’re… we basically have… we’re kind of settling into it. I think the cancellations and the templates were kind of the last… one of the last big pieces.
85 00:08:17.730 ⇒ 00:08:18.250 Mustafa Raja: Yeah.
86 00:08:18.250 ⇒ 00:08:23.169 Sam Roberts: gonna… Like, really feel like this is definitely the right architecture, but I think.
87 00:08:23.170 ⇒ 00:08:23.590 Amber Lin: Okay.
88 00:08:23.590 ⇒ 00:08:26.269 Sam Roberts: We’re kind of tacitly already there, but…
89 00:08:26.270 ⇒ 00:08:27.190 Amber Lin: Cool, okay.
90 00:08:27.190 ⇒ 00:08:28.909 Sam Roberts: Not officially, yeah.
91 00:08:29.320 ⇒ 00:08:35.740 Amber Lin: Yeah, for these two, should I push the start date back? Have we.
92 00:08:35.740 ⇒ 00:08:36.190 Sam Roberts: Startup.
93 00:08:37.250 ⇒ 00:08:38.549 Sam Roberts: No, yeah, push…
94 00:08:39.220 ⇒ 00:08:41.419 Amber Lin: So, next week, or…
95 00:08:41.429 ⇒ 00:08:53.669 Mustafa Raja: So load test routing, this really looks like, something to do when we have, have it deployed on, on our final architecture, right? Yeah, Google.
96 00:08:53.869 ⇒ 00:08:54.579 Mustafa Raja: So…
97 00:08:54.580 ⇒ 00:08:55.190 Casie Aviles: Okay.
98 00:08:56.380 ⇒ 00:08:59.820 Amber Lin: Does that… We need to move it after…
99 00:09:00.040 ⇒ 00:09:09.970 Sam Roberts: Good question. Let me… so routing… yeah, I mean, the load testing… Under concurrent usage…
100 00:09:10.150 ⇒ 00:09:13.570 Sam Roberts: I mean, Monster should be able to handle that, like, it’s not, like, an issue with the architecture.
101 00:09:13.570 ⇒ 00:09:15.789 Mustafa Raja: Yeah, it should be able to, yeah.
102 00:09:15.790 ⇒ 00:09:16.860 Sam Roberts: Yeah, so that’s probably.
103 00:09:16.860 ⇒ 00:09:27.929 Mustafa Raja: Yeah, it’s just… the only question there would be if server is also able to handle that one. I think GCP is… GCP is pretty good, but…
104 00:09:27.930 ⇒ 00:09:29.480 Sam Roberts: I’m not too worried.
105 00:09:30.960 ⇒ 00:09:33.900 Amber Lin: Okay, so…
106 00:09:33.900 ⇒ 00:09:37.790 Casie Aviles: Yeah, we haven’t dived too deep into the GCP part yet.
107 00:09:37.790 ⇒ 00:09:39.690 Sam Roberts: No, no, that was kind of all.
108 00:09:39.800 ⇒ 00:09:41.569 Sam Roberts: Okay, down here, right?
109 00:09:41.990 ⇒ 00:09:42.670 Amber Lin: Yeah.
110 00:09:43.720 ⇒ 00:09:45.529 Amber Lin: those are kinda in the future.
111 00:09:45.530 ⇒ 00:09:52.089 Sam Roberts: Yeah, I mean, this is… if we look at it, it’s part of retrieval and routing accuracy, so I think we just want to make sure that there’s not going to be any kind of weird
112 00:09:52.370 ⇒ 00:09:54.399 Sam Roberts: Weirdness with it, you know?
113 00:09:55.160 ⇒ 00:10:02.820 Sam Roberts: Like, full load testing definitely belongs down here, but, like, we can just make sure that we can pass it a few things and it doesn’t get confused, but…
114 00:10:02.820 ⇒ 00:10:04.210 Amber Lin: Yeah. I’m not too concerned.
115 00:10:04.600 ⇒ 00:10:09.140 Amber Lin: Okay, so I think latest, we need to start on the next week, right?
116 00:10:09.140 ⇒ 00:10:10.319 Sam Roberts: Yeah, that’s fine.
117 00:10:10.880 ⇒ 00:10:11.400 Amber Lin: Okay.
118 00:10:11.400 ⇒ 00:10:18.750 Sam Roberts: We might want to even eventually add another, like, add another real load test down to the GCP stuff, but that’s a few weeks out, so let’s not worry about that.
119 00:10:19.290 ⇒ 00:10:25.480 Amber Lin: Cool, okay. Let’s see…
120 00:10:25.610 ⇒ 00:10:27.309 Sam Roberts: What else we got?
121 00:10:27.310 ⇒ 00:10:31.320 Amber Lin: on anything? Like, fixed performance or correctness issues.
122 00:10:32.940 ⇒ 00:10:37.099 Sam Roberts: That’s kind of a… I mean, a general catch-all, I guess?
123 00:10:37.400 ⇒ 00:10:41.209 Mustafa Raja: Yeah, this looks like, an ongoing thing, you know, as we go.
124 00:10:41.810 ⇒ 00:10:43.510 Sam Roberts: Yeah, I wouldn’t…
125 00:10:43.960 ⇒ 00:10:48.409 Amber Lin: Yeah, I mean, it’s dependent on lots of things, so it’s not really, like, a, you know…
126 00:10:48.760 ⇒ 00:10:56.910 Amber Lin: Alright, and then that’s in progress. What about this? I know, Pranav was trying to work on this.
127 00:10:57.070 ⇒ 00:10:57.590 Amber Lin: Is there.
128 00:10:57.590 ⇒ 00:10:58.200 Sam Roberts: Yeah, did you get a chance?
129 00:10:58.520 ⇒ 00:11:01.610 Sam Roberts: chat with him yet today? Or you chatted with him yesterday, right, Casey?
130 00:11:02.060 ⇒ 00:11:08.469 Casie Aviles: Yeah, I did, I, I think I owe him… Something like the logs.
131 00:11:08.850 ⇒ 00:11:10.759 Amber Lin: But this should be assigned to him.
132 00:11:11.110 ⇒ 00:11:12.050 Sam Roberts: Okay.
133 00:11:12.050 ⇒ 00:11:12.820 Amber Lin: Okay.
134 00:11:13.120 ⇒ 00:11:14.669 Sam Roberts: Yeah, I mean, is this…
135 00:11:16.630 ⇒ 00:11:22.170 Sam Roberts: I mean, I know it’s going to be related to the evaluation data set stuff, too, so I would definitely…
136 00:11:22.470 ⇒ 00:11:23.500 Sam Roberts: At least…
137 00:11:23.980 ⇒ 00:11:29.860 Sam Roberts: like, looping him into what’s going on there, Mustafa would be good. So, like, once we have that data set clean, it’s good to…
138 00:11:30.830 ⇒ 00:11:36.200 Sam Roberts: Like, good to tie those together.
139 00:11:37.810 ⇒ 00:11:38.540 Casie Aviles: Yeah.
140 00:11:39.610 ⇒ 00:11:41.689 Casie Aviles: he defi- he- I told him, like.
141 00:11:41.880 ⇒ 00:11:50.419 Casie Aviles: Or what he told me is he want to… he wanted to take a look at, you know, the sample logs and the exchanges that were…
142 00:11:50.980 ⇒ 00:11:51.800 Casie Aviles: being done.
143 00:11:51.800 ⇒ 00:11:56.069 Mustafa Raja: Does he have access to, the Snowflake theme?
144 00:11:58.300 ⇒ 00:12:00.720 Casie Aviles: I don’t… I don’t think he has any access yet.
145 00:12:01.090 ⇒ 00:12:06.130 Sam Roberts: Yeah, if he doesn’t have access… I mean, we can definitely give him access, but I think, like, the evaluation data set is going to be…
146 00:12:06.960 ⇒ 00:12:12.800 Sam Roberts: like, the main thing. I mean, he can definitely look through it, I don’t have an issue with that. I think that’s probably good, too, but…
147 00:12:13.220 ⇒ 00:12:14.100 Sam Roberts: Like, what we’re about.
148 00:12:14.100 ⇒ 00:12:16.470 Casie Aviles: She has… yeah, the export.
149 00:12:16.750 ⇒ 00:12:18.120 Casie Aviles: That should be good to know.
150 00:12:18.120 ⇒ 00:12:18.860 Sam Roberts: I think so.
151 00:12:21.100 ⇒ 00:12:25.719 Amber Lin: Do you think this is a reasonable deadline for end of week on this task?
152 00:12:26.230 ⇒ 00:12:28.220 Sam Roberts: Probably not, at this point.
153 00:12:28.950 ⇒ 00:12:29.620 Amber Lin: Mmm.
154 00:12:30.220 ⇒ 00:12:34.289 Sam Roberts: I mean, I’m just thinking, like, that’s another 2 days. I mean, maybe it could be, I just don’t know where…
155 00:12:34.720 ⇒ 00:12:36.470 Sam Roberts: Great, how long it’s gonna take with him.
156 00:12:37.170 ⇒ 00:12:38.749 Casie Aviles: Yeah, I’m not too sure where he’s at.
157 00:12:40.130 ⇒ 00:12:40.760 Amber Lin: Okay.
158 00:12:40.760 ⇒ 00:12:41.569 Casie Aviles: Makes sense.
159 00:12:41.570 ⇒ 00:12:46.070 Sam Roberts: I would leave it for now, and if it needs to go. I don’t think anything’s super dependent on that, is there?
160 00:12:46.370 ⇒ 00:12:48.260 Amber Lin: Not really, I mean, except.
161 00:12:48.260 ⇒ 00:12:53.899 Sam Roberts: Yeah, like, there’s definitely, like, testing’s gonna happen, but, like, yeah, we’re getting to… yeah, but we have another week before we really get to, like, this kind of stuff.
162 00:12:53.900 ⇒ 00:13:00.620 Amber Lin: Yeah, go ahead. Seems like we also have this one that needs to start. Is this gonna start this week, or should I move it back to next week?
163 00:13:00.620 ⇒ 00:13:02.020 Sam Roberts: Validate texting.
164 00:13:03.660 ⇒ 00:13:06.099 Amber Lin: This is more of, like, the…
165 00:13:06.100 ⇒ 00:13:06.790 Sam Roberts: Yeah, zip code.
166 00:13:06.790 ⇒ 00:13:09.499 Amber Lin: Difficult, stuff.
167 00:13:10.570 ⇒ 00:13:12.419 Sam Roberts: I mean, I think that’s probably…
168 00:13:16.140 ⇒ 00:13:22.659 Sam Roberts: Sorry, I’m just thinking… Like, is that going to be related to the evaluation spreadsheet, or the evaluation.
169 00:13:22.660 ⇒ 00:13:25.949 Amber Lin: We would just use a dataset to test this.
170 00:13:25.950 ⇒ 00:13:26.720 Sam Roberts: Yeah…
171 00:13:26.720 ⇒ 00:13:35.810 Amber Lin: But a lot of it is, like, we don’t really need the dataset, we can just look at their spreadsheets to confirm if it’s correct. So…
172 00:13:36.090 ⇒ 00:13:36.810 Amber Lin: because this is…
173 00:13:36.810 ⇒ 00:13:38.299 Sam Roberts: Oh, I see what you’re saying, yeah, yeah.
174 00:13:38.300 ⇒ 00:13:41.549 Amber Lin: Create the right cell in the, in the database.
175 00:13:41.550 ⇒ 00:13:45.189 Sam Roberts: Yeah, so we… but I mean, what’s the completion on that, then next… like, mid-next week?
176 00:13:45.920 ⇒ 00:13:51.619 Amber Lin: Yeah, we can move it to next week, because I know we are working on a database this week.
177 00:13:51.620 ⇒ 00:13:54.009 Sam Roberts: Yeah, I would, I would just shift that out to next week.
178 00:13:54.160 ⇒ 00:13:55.170 Amber Lin: Okay.
179 00:13:55.170 ⇒ 00:13:56.060 Sam Roberts: Yeah, perfect.
180 00:13:57.020 ⇒ 00:13:57.810 Amber Lin: Alright.
181 00:13:58.480 ⇒ 00:14:03.760 Amber Lin: It seems like everything that’s in progress, we’re slightly behind.
182 00:14:03.760 ⇒ 00:14:05.070 Sam Roberts: What are we…
183 00:14:05.070 ⇒ 00:14:07.939 Amber Lin: Did shift some stuff back.
184 00:14:08.300 ⇒ 00:14:11.720 Amber Lin: So, like, these… these would need to start…
185 00:14:12.080 ⇒ 00:14:17.359 Amber Lin: Next week, and do you think we can start link, like, implement LengthViews next week?
186 00:14:18.280 ⇒ 00:14:23.560 Amber Lin: This… this is… and these two are what we have for next week.
187 00:14:23.560 ⇒ 00:14:26.249 Sam Roberts: Yeah, I mean, we can definitely start it.
188 00:14:30.800 ⇒ 00:14:33.089 Amber Lin: I think, I think we’re okay with that.
189 00:14:33.090 ⇒ 00:14:46.630 Sam Roberts: I mean, in terms of, like, setting up the evaluation, like, having all next week to do that for actual implementation, you know, even if something like the established evaluation data set goes a little longer, or the core evaluation metrics are a little longer, like, there’s still stuff to set up and, you know.
190 00:14:46.780 ⇒ 00:14:48.310 Sam Roberts: Pipe everything together.
191 00:14:48.550 ⇒ 00:14:49.719 Amber Lin: Yeah, sounds good.
192 00:14:49.940 ⇒ 00:14:59.380 Amber Lin: So… Casey, could you send what Pranav needs? So… and then we’ll ask him tomorrow at stand-up if he was able to do anything.
193 00:15:01.040 ⇒ 00:15:03.769 Casie Aviles: Sure, sure, yeah, I’m exporting it right now.
194 00:15:04.630 ⇒ 00:15:04.960 Sam Roberts: Thank you.
195 00:15:05.430 ⇒ 00:15:10.209 Amber Lin: Okay, and we can talk about cancellations, then?
196 00:15:10.590 ⇒ 00:15:16.059 Amber Lin: If there’s nothing else here. Okay. Mustafa, do you wanna… do you wanna kick it off?
197 00:15:18.290 ⇒ 00:15:20.669 Mustafa Raja: Yeah, so…
198 00:15:24.020 ⇒ 00:15:30.360 Mustafa Raja: So, so I believe, you want to move from, move away from verbatim responses?
199 00:15:32.820 ⇒ 00:15:37.129 Sam Roberts: I think, I think I made it very complicated. It’s, essentially…
200 00:15:37.130 ⇒ 00:15:37.890 Amber Lin: It’s…
201 00:15:38.390 ⇒ 00:15:51.450 Amber Lin: I’m worried that if we add more and more in the future, it will… it will break the verbatim response, but it seems like that’s not gonna happen. Like, I think that, based on what you guys said, the agent can handle more stuff.
202 00:15:51.580 ⇒ 00:15:52.550 Amber Lin: So…
203 00:15:52.550 ⇒ 00:15:52.990 Mustafa Raja: Oh, let’s.
204 00:15:52.990 ⇒ 00:15:54.990 Sam Roberts: just ignore what I said there.
205 00:15:54.990 ⇒ 00:15:58.030 Amber Lin: I think my question was more of…
206 00:15:58.260 ⇒ 00:16:00.950 Amber Lin: How do we not dump a…
207 00:16:00.950 ⇒ 00:16:02.280 Sam Roberts: That’s what I’m worried about, yeah.
208 00:16:02.280 ⇒ 00:16:07.390 Amber Lin: 5-page-long text onto the agent when they’re handling a customer that wants to cancel.
209 00:16:08.890 ⇒ 00:16:09.460 Sam Roberts: I mean.
210 00:16:09.460 ⇒ 00:16:10.073 Mustafa Raja: Okay, so…
211 00:16:10.380 ⇒ 00:16:11.579 Sam Roberts: Yeah, go ahead, go ahead and stop them.
212 00:16:11.690 ⇒ 00:16:22.839 Mustafa Raja: Yeah, so, so if the question, would be more related to, or more, explanatory in terms of why
213 00:16:22.970 ⇒ 00:16:29.919 Mustafa Raja: Customer, customer wants to cancel, like, if customer wants to cancel because they are moving.
214 00:16:30.060 ⇒ 00:16:34.640 Mustafa Raja: Then we can, select only the script.
215 00:16:34.750 ⇒ 00:16:36.550 Mustafa Raja: For the moving part.
216 00:16:37.160 ⇒ 00:16:43.379 Mustafa Raja: And this is… and this is an example that I have attached within the screenshot in the thread.
217 00:16:43.510 ⇒ 00:16:55.470 Mustafa Raja: So, it really depends on the question of what, to what is going to be returned, you know? So, we can, we can select parts of the whole flow.
218 00:16:56.050 ⇒ 00:16:57.679 Mustafa Raja: We are able to do that.
219 00:16:58.270 ⇒ 00:17:10.169 Amber Lin: I see. We can think about the scenarios of what would they want to do. I mean, they… they would say… I looked at the questions they usually ask, things like, how do I…
220 00:17:10.440 ⇒ 00:17:29.219 Amber Lin: how do I do these things? I think our main goal is to, one, provide them with how to do things, which is, like, which is pretty standard of the SOPs, but more of they’re not using safe tactics, and Yvette really wants them to use safe tactics.
221 00:17:29.360 ⇒ 00:17:34.460 Amber Lin: And… Like, I guess the script is only to help them.
222 00:17:34.460 ⇒ 00:17:34.890 Sam Roberts: Right.
223 00:17:34.890 ⇒ 00:17:38.730 Amber Lin: That, because they… they might get…
224 00:17:38.880 ⇒ 00:17:44.830 Amber Lin: Like, confused or scared when they’re handling a canceling customer, but it’s…
225 00:17:45.370 ⇒ 00:17:50.299 Amber Lin: like, these are nice-to-haves, but I just don’t know if…
226 00:17:50.860 ⇒ 00:17:53.300 Amber Lin: Giving them everything, or even just.
227 00:17:53.300 ⇒ 00:17:53.840 Sam Roberts: Yeah.
228 00:17:53.840 ⇒ 00:17:59.680 Amber Lin: section really tells them all the safe tactics that they need to say.
229 00:18:00.430 ⇒ 00:18:01.100 Sam Roberts: Right.
230 00:18:01.320 ⇒ 00:18:02.510 Amber Lin: they have…
231 00:18:02.980 ⇒ 00:18:18.700 Amber Lin: Because even in moving, it’s, okay, where are they moving? Depending on, like, what area are they moving to? They have different options to offer, right? If it’s in their service area, or if it’s outside, or…
232 00:18:19.230 ⇒ 00:18:36.459 Amber Lin: or, outside of state where none of their, like, brother branches are at, then, like, they have different ways to handle it, and then ultimately, they also have some stuff they can offer, and ideally, if we can give them, like… if…
233 00:18:37.300 ⇒ 00:18:46.349 Amber Lin: Like, we don’t have to generate scripts, but they will give us the scripts that will be helpful, but essentially, it’s how do we…
234 00:18:47.970 ⇒ 00:18:53.889 Amber Lin: I just… do you… I just don’t know if, giving them everything at the same time.
235 00:18:54.300 ⇒ 00:18:55.520 Sam Roberts: No, I think you’re absolutely right.
236 00:18:55.520 ⇒ 00:18:56.799 Amber Lin: Good idea.
237 00:18:56.800 ⇒ 00:18:57.490 Casie Aviles: I mean…
238 00:18:57.490 ⇒ 00:18:58.890 Sam Roberts: split. Go ahead.
239 00:18:59.180 ⇒ 00:19:03.159 Casie Aviles: Oh, yeah, maybe we could chunk it, I guess? I mean…
240 00:19:04.820 ⇒ 00:19:05.979 Casie Aviles: In a way, but…
241 00:19:06.090 ⇒ 00:19:12.270 Casie Aviles: is it still important that we keep it verbatim then, right? I believe it is, then…
242 00:19:12.270 ⇒ 00:19:15.150 Sam Roberts: Well, certain parts of it seem like they’re important for verbatim, right?
243 00:19:15.370 ⇒ 00:19:21.419 Casie Aviles: Yeah, and just… we basically just send it out in chunks, and then… the CSRs would ask?
244 00:19:21.610 ⇒ 00:19:24.539 Casie Aviles: You know, for, for more.
245 00:19:25.000 ⇒ 00:19:29.360 Casie Aviles: And then, that’s probably what I was… Thinking, but… Do you think…
246 00:19:29.360 ⇒ 00:19:45.100 Amber Lin: I think it’s good to have, like what we did with the attributes, I think this is a good use of follow-up questions. Like, we can say, I want to… the customer is gonna move, and then we say, hey, this is what you can say, and these are the three options. Are they…
247 00:19:45.100 ⇒ 00:19:57.679 Amber Lin: moving inside state, outside state, in another branch, and then based on what they say, we can give them the next part, and then add another follow-up question. It’s like, hey, have you
248 00:19:57.690 ⇒ 00:20:04.079 Amber Lin: Have you said these safe tactics? These are some safe tactics you can say, and then…
249 00:20:04.250 ⇒ 00:20:17.449 Amber Lin: And then… and then asked another question. It was like, hey, are… is the customer still wanting to cancel? If… if not, then they directed to, like, the procedure of doing whatever the safe tax says.
250 00:20:17.450 ⇒ 00:20:24.859 Amber Lin: Or if the customer still wants to cancel, then we go to step 3, and we… we tell them, like, hey, this is,
251 00:20:25.710 ⇒ 00:20:34.310 Amber Lin: This is step… step 3 of, okay, give them reward points, and then lastly, do the cancellations.
252 00:20:34.580 ⇒ 00:20:38.230 Amber Lin: I was just wondering, do you think that’s possible?
253 00:20:38.740 ⇒ 00:20:44.409 Sam Roberts: So, like, the real thing here, like, the ideal use, or the ideal,
254 00:20:45.680 ⇒ 00:20:51.699 Sam Roberts: state, I guess, is that this is basically a giant flow chart that we want Andy to help them work their way
255 00:20:54.320 ⇒ 00:20:57.120 Sam Roberts: Not just a giant piece of text that we want to return.
256 00:20:57.120 ⇒ 00:21:00.459 Amber Lin: Yeah, flowchart is… is the right word.
257 00:21:00.670 ⇒ 00:21:04.610 Sam Roberts: Right, so… You know, in a…
258 00:21:06.180 ⇒ 00:21:11.790 Sam Roberts: How many of these cancellations are there? Like, different flows? Is it per… it’s, like, one per department or something like that?
259 00:21:11.790 ⇒ 00:21:19.610 Amber Lin: No, we’re… I’m actually trying to work with them to consolidate the department, so ultimately, there will probably be, like.
260 00:21:20.080 ⇒ 00:21:28.919 Amber Lin: five main reasons, because it only matters if they… if they’re… if the save tactics are different, so…
261 00:21:29.310 ⇒ 00:21:35.259 Sam Roberts: Okay. Yeah, I mean… I don’t love the idea of making, like, a custom flow.
262 00:21:35.820 ⇒ 00:21:38.970 Sam Roberts: you know, like, we… we want to kind of keep the AI
263 00:21:39.940 ⇒ 00:21:44.889 Sam Roberts: Helping more than just programming the, like, the flowchart in.
264 00:21:45.050 ⇒ 00:21:52.020 Sam Roberts: But if it’s something where, like, we are gonna consolidate, and this is, like, an important thing to, like, have them walk through, like, maybe we need to build another…
265 00:21:52.230 ⇒ 00:21:54.610 Sam Roberts: Like.
266 00:21:54.610 ⇒ 00:21:56.610 Amber Lin: an interface, maybe? Just for.
267 00:21:56.610 ⇒ 00:22:15.270 Sam Roberts: But, like, no, like, so in Monster, you can have agents, tools, and workflows, and we’re, like, we’re kind of starting to use a few different ones of those. So, like, the tool calls are, like, accessing the database, and, like, the agents are, like, the prompted things, and then the workflows are kind of all of these things plugged together. So I’m wondering if maybe we need to have another
268 00:22:16.590 ⇒ 00:22:32.110 Sam Roberts: and I don’t… I don’t love this idea, because it’s a little more, like, ad hoc, but if it’s something that’s important, we might want to just do it, where it’s, like, a cancellation workflow, and it knows how to read this document and send people in the right… like, you know, it prompts them to say, like.
269 00:22:32.830 ⇒ 00:22:38.620 Sam Roberts: which answer did I give, A, B, C, or D? And then it jumps to that part for them. Yeah.
270 00:22:39.170 ⇒ 00:22:44.260 Sam Roberts: I… You know, we’re getting a little more into the, like.
271 00:22:45.110 ⇒ 00:22:52.879 Sam Roberts: it’s gonna matter how that document is structured, and how it’s, like… So, if things change, that means that Andy would have to change on the back end, I think, right now?
272 00:22:55.070 ⇒ 00:22:58.700 Sam Roberts: So if we’re going to, like,
273 00:22:59.560 ⇒ 00:23:07.260 Sam Roberts: you know, stabilize these a little bit more, and consolidate, I think that’s probably a good time to then say, okay, we can have a, like.
274 00:23:08.460 ⇒ 00:23:14.239 Sam Roberts: A flow that this… You know, bot can follow along and help the agents.
275 00:23:17.170 ⇒ 00:23:22.790 Sam Roberts: I don’t know, I mean… Gustava, Casey, what do you guys think about something like that? Is that, like…
276 00:23:23.140 ⇒ 00:23:28.429 Sam Roberts: a bigger… bill than is worth it? Like, is that a higher…
277 00:23:28.550 ⇒ 00:23:33.099 Sam Roberts: A bigger lift than, like, what we could maybe do with just some chunking and…
278 00:23:33.370 ⇒ 00:23:37.960 Sam Roberts: Yeah, I could figure out the right spots, or do you think we should add a little more of our own logic in there?
279 00:23:38.220 ⇒ 00:23:41.129 Amber Lin: Or should we just have them, like, when we…
280 00:23:41.330 ⇒ 00:23:47.459 Amber Lin: reference to this document, because I know it meant we’re trying to do, like, show the source.
281 00:23:47.720 ⇒ 00:23:48.300 Sam Roberts: Yeah.
282 00:23:48.300 ⇒ 00:23:52.650 Amber Lin: First case, we can just say, hey, Click on this document.
283 00:23:53.080 ⇒ 00:24:01.130 Sam Roberts: That, I mean, that’s… that’s the, like, easiest thing, I think, is, like, identifying what the cancellation’s for and sending them there. The other extreme…
284 00:24:01.130 ⇒ 00:24:08.769 Amber Lin: know about certain things, they can ask Andy, and then we can… that will just be the same as the central dog, it’ll just be brag.
285 00:24:09.730 ⇒ 00:24:15.419 Sam Roberts: Yeah, that might be the… Best thing to start with, and if we find that that’s…
286 00:24:15.530 ⇒ 00:24:24.429 Sam Roberts: not a good, like, user experience, then we go and say, okay, well, we’re going to build a little, you know, state machine, basically, that can bounce through the
287 00:24:24.920 ⇒ 00:24:37.919 Sam Roberts: you know, if you’re here, and they say this, go here, and then it just returns the script to them. But I think in the first pass, like, that’s what you’ve just outlined is probably a better, more efficient use of time and energy.
288 00:24:38.420 ⇒ 00:24:39.180 Amber Lin: Okay.
289 00:24:40.180 ⇒ 00:24:41.639 Casie Aviles: Yeah, because… So maybe that…
290 00:24:41.800 ⇒ 00:24:42.640 Sam Roberts: Go ahead.
291 00:24:42.640 ⇒ 00:24:47.430 Casie Aviles: Yeah, I feel like we’re gonna build a lot for just the cancellation.
292 00:24:47.430 ⇒ 00:24:49.219 Sam Roberts: That’s what I feel like, too, yeah.
293 00:24:49.940 ⇒ 00:25:07.969 Amber Lin: Yeah, and it’s, like, I don’t think it’s worth it that much, because I think this is… because this relates more to revenue than anything, because this is just money going out the… out the door, so that’s why they care so much, but I really don’t see any other part.
294 00:25:08.080 ⇒ 00:25:09.080 Amber Lin: Yes.
295 00:25:09.080 ⇒ 00:25:09.780 Sam Roberts: Yeah.
296 00:25:09.940 ⇒ 00:25:16.690 Amber Lin: that needs this… this level of, hey, you must save these, save tactics. You’re absolutely right.
297 00:25:16.900 ⇒ 00:25:17.450 Amber Lin: Family.
298 00:25:17.450 ⇒ 00:25:37.249 Sam Roberts: And I think… I think that’s good insight, though, too. Like, if it becomes… where this, like, return the document, have them click around the document on their own is not good enough, and they want something built a little more, you know, then because there’s losing revenue, like, there could be a justification for something more complex, but I don’t think there is right now. I think you’re right.
299 00:25:39.410 ⇒ 00:25:41.990 Sam Roberts: But I would say let’s try that, so we can…
300 00:25:43.580 ⇒ 00:25:45.170 Sam Roberts: Well, that… ugh, I don’t know.
301 00:25:48.030 ⇒ 00:25:52.120 Sam Roberts: Maybe we return the initial part of the cancellation with a link to it and everything?
302 00:25:53.490 ⇒ 00:25:58.230 Amber Lin: Yeah, we can, like, we can just return, like, hey, these are the…
303 00:25:59.220 ⇒ 00:26:15.420 Amber Lin: We can… maybe, like, a standard script of, hey, these are the, common reasons, these are some common safe tactics you can use, just like a high-level thing they can reference, and then say, hey, if you need help with scripts.
304 00:26:15.560 ⇒ 00:26:21.719 Amber Lin: please click this link. And I can… I can tune that message with… with Yvette and Janiece.
305 00:26:21.720 ⇒ 00:26:22.610 Sam Roberts: Yeah.
306 00:26:22.610 ⇒ 00:26:33.170 Amber Lin: But as long as you think this is a possible workflow, we can just link the… we can just link a doc, because they’re spending so much time, like, working on this doc anyways.
307 00:26:33.170 ⇒ 00:26:33.880 Sam Roberts: Yeah.
308 00:26:33.880 ⇒ 00:26:35.729 Amber Lin: I might as well just…
309 00:26:35.730 ⇒ 00:26:36.380 Sam Roberts: Okay.
310 00:26:36.910 ⇒ 00:26:37.899 Amber Lin: Give it to them.
311 00:26:38.630 ⇒ 00:26:45.469 Sam Roberts: Yeah, I think that’s unfortunately probably the best way to do it. Or not the best, but, like, the best of the options for everything.
312 00:26:47.770 ⇒ 00:26:48.390 Amber Lin: Cool.
313 00:26:49.140 ⇒ 00:26:53.480 Sam Roberts: Yeah, and yeah, once I saw that, like, massive wall of text, I was like, oh, this isn’t gonna be quite…
314 00:26:53.870 ⇒ 00:26:54.940 Sam Roberts: Helpful.
315 00:26:57.160 ⇒ 00:26:58.050 Amber Lin: Okay.
316 00:26:58.220 ⇒ 00:27:00.939 Amber Lin: Mustafa, what do you think? So, like…
317 00:27:00.940 ⇒ 00:27:01.310 Sam Roberts: Yeah.
318 00:27:01.310 ⇒ 00:27:04.279 Mustafa Raja: Yeah, I think the same. I think the same.
319 00:27:08.910 ⇒ 00:27:13.180 Mustafa Raja: Yeah, the textage is too much, and then…
320 00:27:14.280 ⇒ 00:27:24.189 Mustafa Raja: The other… the other thing is we are… we’re back… we are back with the same problem with the RAG, you know? It’s not going to exactly say… say the same tactics or something.
321 00:27:25.100 ⇒ 00:27:25.430 Sam Roberts: Yeah.
322 00:27:25.430 ⇒ 00:27:30.369 Amber Lin: So, it might just be, the smarter thing to link to.
323 00:27:30.620 ⇒ 00:27:32.129 Mustafa Raja: The document.
324 00:27:32.480 ⇒ 00:27:34.560 Mustafa Raja: The smarter and easier.
325 00:27:34.940 ⇒ 00:27:40.010 Sam Roberts: Yeah. Yeah, and I think… I think, like I said, if it becomes a problem.
326 00:27:40.400 ⇒ 00:27:48.370 Sam Roberts: And they want something a little more, you know, chat interface-y, then we can brainstorm some other ways to potentially do that where, you know.
327 00:27:48.560 ⇒ 00:27:57.499 Sam Roberts: maybe we ingest the cancellation document and have a new thing that the AI can traverse more easily, or some… you know, there’s other ways we can figure it out, but not worth the effort right now, I think.
328 00:27:57.500 ⇒ 00:28:01.540 Amber Lin: Yeah, that’s just kind of very scratchy. Okay, and then…
329 00:28:02.200 ⇒ 00:28:17.349 Amber Lin: we’ll still have RAG answer some of the questions, because it sounds pretty easy to grab, like, oh, get me the safe tactics for this, and I think RAG should handle that if we put the headers properly. So I think that’s a good enough solution for now.
330 00:28:17.720 ⇒ 00:28:23.879 Sam Roberts: Yeah, yeah, if they’re asking for specific parts of it, that’s great. We can still do the retrieval and return verbatim smaller parts, yeah.
331 00:28:24.490 ⇒ 00:28:31.710 Amber Lin: Yeah, verbatim, like, if they ask, maybe, like, a keyword for script, and then we just send them, like, the…
332 00:28:31.710 ⇒ 00:28:32.370 Sam Roberts: Yeah.
333 00:28:32.370 ⇒ 00:28:43.060 Mustafa Raja: For now, the keyword is cancellations, so let me know, which keyword would you want, for the router to verbatim?
334 00:28:43.060 ⇒ 00:28:49.699 Amber Lin: Probably, like, if they ever mention cancellations, give them this link, like, regardless, so they…
335 00:28:49.700 ⇒ 00:28:50.490 Sam Roberts: Yeah.
336 00:28:50.740 ⇒ 00:28:51.669 Amber Lin: Go read it.
337 00:28:51.670 ⇒ 00:28:54.639 Mustafa Raja: Like, it’s not ideal. I know they want to…
338 00:28:54.640 ⇒ 00:29:00.940 Amber Lin: not have the folks read docs and just use Andy, but I think that’s too much for us to handle.
339 00:29:01.180 ⇒ 00:29:06.870 Sam Roberts: Yeah, I think right now it’s just, it’s a lot for what’s not maybe worth…
340 00:29:06.870 ⇒ 00:29:07.450 Amber Lin: Okay.
341 00:29:07.630 ⇒ 00:29:12.190 Sam Roberts: worth the effort right now. If it turns out to be, like, it’s definitely something we can look into more.
342 00:29:12.190 ⇒ 00:29:20.320 Amber Lin: Yeah, okay. Sounds good. That was my main worry, and then… Anything else?
343 00:29:21.510 ⇒ 00:29:31.850 Mustafa Raja: Yeah, so I would want to reiterate on what we just discussed. So, I will be updating the verbatim, agent to rather,
344 00:29:32.080 ⇒ 00:29:38.940 Mustafa Raja: Send some, quick save tactics, save tactics, and send in the link, right?
345 00:29:40.560 ⇒ 00:29:46.240 Amber Lin: Yes, you might need… Actually…
346 00:29:48.640 ⇒ 00:29:59.799 Amber Lin: Let me check on which link to add, but, like, the save tactics are right now not directly in the central dock yet, so…
347 00:30:00.370 ⇒ 00:30:04.710 Amber Lin: could we hold off on that until I talk to them? Like, maybe?
348 00:30:04.710 ⇒ 00:30:05.790 Mustafa Raja: Yeah, yeah, we can.
349 00:30:06.100 ⇒ 00:30:06.940 Mustafa Raja: Yeah.
350 00:30:07.690 ⇒ 00:30:17.020 Amber Lin: Yeah. And then do we have a… do we have an absolute keyword that we want? The one mentioned, we would want to route to this agent.
351 00:30:17.960 ⇒ 00:30:20.029 Mustafa Raja: Or is it just going to be cancellation?
352 00:30:20.600 ⇒ 00:30:22.210 Amber Lin: I think it’s just cancellation.
353 00:30:22.560 ⇒ 00:30:23.360 Sam Roberts: Yeah, I’ll start with that.
354 00:30:23.360 ⇒ 00:30:27.210 Mustafa Raja: So, so whenever the word cancellation would be…
355 00:30:27.450 ⇒ 00:30:32.800 Mustafa Raja: Mentioned, the router would route to this specification. Yeah.
356 00:30:32.970 ⇒ 00:30:35.810 Amber Lin: Yeah, I was like, if… yeah.
357 00:30:38.420 ⇒ 00:30:49.829 Casie Aviles: Jet, I guess I have a quick question regarding that. What about, like… I don’t want to complicate it, but they have questions, for example,
358 00:30:50.090 ⇒ 00:30:52.239 Casie Aviles: Customer wants to cancel.
359 00:30:52.490 ⇒ 00:30:56.249 Casie Aviles: And then they state some reason. Do we consider that as well, or…
360 00:30:56.430 ⇒ 00:30:57.599 Casie Aviles: Not for now.
361 00:30:59.950 ⇒ 00:31:11.839 Amber Lin: I think so. If they mention anything related to cancel, we’ll try to answer it with Andy, with Rag, and whatever, but we should also send a link, because that’s a more…
362 00:31:12.030 ⇒ 00:31:13.440 Amber Lin: like, robust.
363 00:31:14.250 ⇒ 00:31:14.800 Amber Lin: one…
364 00:31:14.800 ⇒ 00:31:15.190 Casie Aviles: We can’.
365 00:31:15.390 ⇒ 00:31:21.389 Amber Lin: I don’t know how the prompting would work. We can instruct it to try and be more…
366 00:31:22.130 ⇒ 00:31:29.250 Amber Lin: Like, detailed and ask follow-up questions when asked about cancellations, but ultimately, like.
367 00:31:29.380 ⇒ 00:31:38.879 Amber Lin: Like, we can… if they want scripts, we can refer them to the doc, but if they just want save tactics, or how to do things, I think RAG is enough.
368 00:31:39.550 ⇒ 00:31:41.479 Amber Lin: I hope so, we should test it.
369 00:31:41.610 ⇒ 00:31:42.530 Amber Lin: So…
370 00:31:44.060 ⇒ 00:31:44.640 Casie Aviles: Okay.
371 00:31:51.620 ⇒ 00:31:52.449 Sam Roberts: Any other…
372 00:31:54.380 ⇒ 00:32:12.509 Mustafa Raja: Yeah, just quick clarification that the RAG Andy would be… is a different agent, and then this verbatim response agent is a different agent, right? So, if they ever mention cancellations, the RAG Andy isn’t going to answer anything, it’s just going to be
373 00:32:12.580 ⇒ 00:32:17.350 Mustafa Raja: The verbatim response, or whatever we instruct this verbatim agent to do.
374 00:32:19.000 ⇒ 00:32:27.749 Sam Roberts: Yeah, so I think that if they mention cancellations, we should make sure to include that link, and then depending on maybe what else is in that message, the AI can…
375 00:32:28.280 ⇒ 00:32:28.930 Sam Roberts: do…
376 00:32:28.930 ⇒ 00:32:29.400 Amber Lin: Yeah.
377 00:32:29.400 ⇒ 00:32:33.790 Sam Roberts: A verbatim response of maybe some save tactics, but not the whole cancellation.
378 00:32:35.450 ⇒ 00:32:36.590 Mustafa Raja: I just wanted to confirm that.
379 00:32:38.880 ⇒ 00:32:41.370 Mustafa Raja: I just wanted to confirm that, yeah, thank you.
380 00:32:41.370 ⇒ 00:32:43.080 Sam Roberts: Yeah, I think, I think that’s still the right…
381 00:32:43.180 ⇒ 00:32:49.579 Sam Roberts: Like, architecture where it’ll go to a separate sub-agent like that, but that sub-agent will just need to make sure to include that link
382 00:32:49.730 ⇒ 00:32:56.610 Sam Roberts: And then… Return… Whatever it thinks is the most relevant chunk, maybe?
383 00:32:56.770 ⇒ 00:32:57.540 Mustafa Raja: Okay.
384 00:32:57.870 ⇒ 00:32:58.400 Mustafa Raja: Yeah.
385 00:32:58.400 ⇒ 00:33:07.220 Sam Roberts: Based on if there’s other information. Like, if they mention a reason, maybe it says, like, oh, if that’s the reason, you know, try this. If you want to look up a little bit more, click this link, kind of thing.
386 00:33:07.860 ⇒ 00:33:08.490 Mustafa Raja: Yep.
387 00:33:11.060 ⇒ 00:33:11.470 Amber Lin: Yeah.
388 00:33:11.470 ⇒ 00:33:11.890 Sam Roberts: All right.
389 00:33:11.890 ⇒ 00:33:31.459 Mustafa Raja: Also, the verbatim thing is only the cancellation, right? So, no other verbatim responses. So, because right now, in the documentation, or in the prompt, I have called this documentation as verbatim responses also, so I’ll just remove that then, right?
390 00:33:31.460 ⇒ 00:33:38.629 Amber Lin: Yeah, I think call lists is… we can remove that. Is this so in the, like, the main prompt? Can we just have.
391 00:33:38.630 ⇒ 00:33:39.020 Mustafa Raja: Perfect.
392 00:33:39.020 ⇒ 00:33:46.029 Amber Lin: It’s a dock, so it has… it has separate contacts, so it doesn’t need to crowd the central prompt.
393 00:33:47.250 ⇒ 00:33:53.180 Mustafa Raja: Yeah, I think, I think all this, mentioned in the… what’s it called? In the prompt.
394 00:33:53.720 ⇒ 00:33:56.300 Mustafa Raja: prompt design, LangFuse, we can integrate within that.
395 00:33:58.050 ⇒ 00:33:59.960 Sam Roberts: Which part is this? I’m sorry, I’m gonna pull on.
396 00:34:01.630 ⇒ 00:34:07.000 Mustafa Raja: So the prompt also has, what’s it called?
397 00:34:07.180 ⇒ 00:34:13.099 Mustafa Raja: A verbatim response for, call list documentation.
398 00:34:13.460 ⇒ 00:34:14.760 Sam Roberts: Oh, oh.
399 00:34:15.100 ⇒ 00:34:16.090 Mustafa Raja: Yeah.
400 00:34:16.090 ⇒ 00:34:24.249 Amber Lin: I think that’s… we can remove that if… if that’s… I don’t think that’s a big issue. My question was, is there still, like, the whole…
401 00:34:24.520 ⇒ 00:34:27.059 Amber Lin: Text or cancellations in our prompt.
402 00:34:27.659 ⇒ 00:34:29.459 Mustafa Raja: Yeah, I think yes. For Annie.
403 00:34:29.460 ⇒ 00:34:29.899 Sam Roberts: Oh, that is.
404 00:34:29.909 ⇒ 00:34:32.709 Mustafa Raja: hosted on NH10, it is still there.
405 00:34:33.329 ⇒ 00:34:34.539 Casie Aviles: Oh, buffalo Master.
406 00:34:34.540 ⇒ 00:34:37.250 Sam Roberts: And the new one, it’s gonna… yeah, no, that’s part of this whole sub-Asian.
407 00:34:37.250 ⇒ 00:34:40.690 Amber Lin: Okay, okay, great, great. That’s essentially what I was asking.
408 00:34:41.050 ⇒ 00:34:46.029 Amber Lin: I’ll give you the document to, like, as context.
409 00:34:46.380 ⇒ 00:34:48.210 Mustafa Raja: Yep.
410 00:34:51.790 ⇒ 00:34:54.460 Amber Lin: And the insect sprouts…
411 00:35:10.680 ⇒ 00:35:11.540 Amber Lin: Cool.
412 00:35:19.840 ⇒ 00:35:20.770 Amber Lin: Okay.
413 00:35:21.030 ⇒ 00:35:22.860 Amber Lin: Alright,
414 00:35:24.440 ⇒ 00:35:30.920 Amber Lin: I would like us to work a little bit on the triage issues, because I don’t think we’ve touched them.
415 00:35:31.430 ⇒ 00:35:32.160 Sam Roberts: Hmm.
416 00:35:32.160 ⇒ 00:35:33.160 Amber Lin: at all.
417 00:35:34.130 ⇒ 00:35:40.219 Casie Aviles: Yeah, there were a few triage tickets that I shaved off last week, but there’s still a lot.
418 00:35:40.220 ⇒ 00:35:40.860 Amber Lin: Awesome.
419 00:35:42.520 ⇒ 00:35:47.600 Casie Aviles: Yeah, I think I got around 3 or 5 triage tickets, 3 to 5.
420 00:35:47.600 ⇒ 00:35:48.180 Amber Lin: -Oh.
421 00:35:49.370 ⇒ 00:35:49.960 Casie Aviles: Cool.
422 00:35:50.070 ⇒ 00:35:51.430 Casie Aviles: But there’s still a lot.
423 00:35:51.860 ⇒ 00:35:55.020 Amber Lin: Yeah, there’s a… we’ll work through them slowly.
424 00:35:55.750 ⇒ 00:36:05.079 Amber Lin: I think, Sam, you can feel free to hop off and, like, we can use the last 20-ish minutes or so to work on triage.
425 00:36:06.330 ⇒ 00:36:08.220 Sam Roberts: Alright, cool. Yeah, let me know if you need anything else.
426 00:36:08.340 ⇒ 00:36:09.490 Amber Lin: Okay. Sounds like.
427 00:36:09.760 ⇒ 00:36:10.760 Sam Roberts: Alright, thanks, y’all.
428 00:36:11.010 ⇒ 00:36:12.880 Amber Lin: Thanks. Bye.
429 00:36:14.670 ⇒ 00:36:15.590 Amber Lin: Okay.
430 00:36:17.620 ⇒ 00:36:19.739 Amber Lin: Working session.
431 00:36:19.980 ⇒ 00:36:26.320 Amber Lin: Should we stay online if there’s any questions, or do you guys just wanna hang out, or…
432 00:36:26.840 ⇒ 00:36:28.200 Amber Lin: Take it offline.
433 00:36:29.740 ⇒ 00:36:32.769 Mustafa Raja: I’m timed to be, in the meeting.
434 00:36:33.930 ⇒ 00:36:34.670 Amber Lin: Okay.
435 00:36:34.940 ⇒ 00:36:37.940 Amber Lin: Cool. Just, like, 25 minutes or so, let’s work.
436 00:36:37.940 ⇒ 00:36:38.830 Mustafa Raja: Yeah.
437 00:36:39.530 ⇒ 00:36:45.439 Amber Lin: Do you have… Mustafa, do you have any triage tickets, or Casey, would you assign some to Mustafa if you need help?
438 00:36:46.830 ⇒ 00:36:47.420 Mustafa Raja: Yes, sir.
439 00:36:48.920 ⇒ 00:36:50.540 Casie Aviles: Let’s see…
440 00:36:54.970 ⇒ 00:36:56.080 Casie Aviles: Yeah, I’ve… I see…
441 00:36:56.080 ⇒ 00:36:58.019 Mustafa Raja: That there are some assigned to me.
442 00:37:10.810 ⇒ 00:37:11.900 Casie Aviles: Me too.
443 00:37:12.090 ⇒ 00:37:13.969 Casie Aviles: Yeah, I’m just gonna go and share.
444 00:37:23.900 ⇒ 00:37:28.830 Casie Aviles: Okay, yeah, it looks like you have some already on your plate as well.
445 00:37:28.830 ⇒ 00:37:29.570 Mustafa Raja: Yep.
446 00:37:36.230 ⇒ 00:37:38.930 Casie Aviles: We’ll just go through… Whichever…
447 00:37:48.810 ⇒ 00:37:49.530 Mustafa Raja: Yeah.
448 00:37:52.920 ⇒ 00:37:56.449 Mustafa Raja: I’m taking a look at the ones that are assigned to me.
449 00:45:02.410 ⇒ 00:45:07.240 Mustafa Raja: Also, Amber, we now have, the…
450 00:45:07.400 ⇒ 00:45:17.929 Mustafa Raja: what’s it called? The service account to pull in data into BigQuery. So, I saw that the other day, you mentioned,
451 00:45:18.210 ⇒ 00:45:20.519 Mustafa Raja: a Google Drive link.
452 00:45:20.800 ⇒ 00:45:23.879 Mustafa Raja: Can you mention that again? And also…
453 00:45:24.030 ⇒ 00:45:31.649 Mustafa Raja: I saw that, it had some PDFs in it, so how do you want to handle that?
454 00:45:32.640 ⇒ 00:45:35.259 Mustafa Raja: Do we want these in BigQuery or somewhere?
455 00:45:36.000 ⇒ 00:45:44.969 Amber Lin: Oh, I remember, okay, so putting the… Google Drive, Docs, and BigQuery.
456 00:45:45.170 ⇒ 00:45:46.389 Mustafa Raja: Yeah.
457 00:45:47.230 ⇒ 00:45:53.350 Amber Lin: I think we’re safe to skip the PDFs. It’s mainly the spreadsheets that we do care about.
458 00:45:53.970 ⇒ 00:46:00.080 Mustafa Raja: Okay. Okay, I’ll, I’ll have the spreadsheets, each tab as a, as a table.
459 00:46:00.210 ⇒ 00:46:05.339 Mustafa Raja: In BigQuery, and then… and then let me know if… if you need anything more.
460 00:46:05.710 ⇒ 00:46:06.979 Amber Lin: I’ll be able to help.
461 00:46:07.480 ⇒ 00:46:12.429 Amber Lin: Yeah, awesome. Do you need… do you have a ticket for that? Want to make sure…
462 00:46:12.430 ⇒ 00:46:16.210 Mustafa Raja: Oh, yeah, yeah, let’s, let’s, let’s, let’s create a ticket for that.
463 00:46:16.210 ⇒ 00:46:18.929 Amber Lin: Yeah, I’ll do that.
464 00:46:19.500 ⇒ 00:46:27.529 Amber Lin: This is, like, discovery work. Okay, I’ll make another… Well, I’ll make another project.
465 00:46:32.330 ⇒ 00:46:36.950 Mustafa Raja: And then you’ll be, querying with the BigQuery
466 00:46:37.380 ⇒ 00:46:46.959 Mustafa Raja: Within the BigQuery, query, interface, or do you need… do you need something else?
467 00:46:47.650 ⇒ 00:47:07.470 Amber Lin: I’ll… I’ll probably explore it in BigQuery. I mainly… I did most of the analysis already. I think this is just so for future use. I downloaded the data and did the analysis in cursor. So, I think it’s just in the future, we may need to reference those tables, that’s all.
468 00:47:08.220 ⇒ 00:47:08.850 Mustafa Raja: Okay.
469 00:47:09.620 ⇒ 00:47:19.389 Amber Lin: Yeah, let me add… Upload… Doing sales data too big.
470 00:47:19.610 ⇒ 00:47:22.649 Amber Lin: Very… I’ll assign it to you.
471 00:47:27.610 ⇒ 00:47:28.720 Amber Lin: Cool. Okay.
472 00:49:08.930 ⇒ 00:49:16.190 Casie Aviles: Oh, by the way, Amber, the messages, the daily messages are already being sent by Andy, so…
473 00:49:16.190 ⇒ 00:49:17.980 Amber Lin: Yeah, that’s… that’s okay.
474 00:49:18.120 ⇒ 00:49:22.170 Amber Lin: I forgot to give it to them, but I hear no complaints, so…
475 00:49:22.460 ⇒ 00:49:24.350 Casie Aviles: I’ll set this to done, then.
476 00:49:24.350 ⇒ 00:49:25.579 Amber Lin: Okay, thank you.
477 00:50:17.760 ⇒ 00:50:24.230 Mustafa Raja: So, does the database agent has, universal inspectors?
478 00:50:27.190 ⇒ 00:50:27.910 Casie Aviles: Sorry?
479 00:50:28.680 ⇒ 00:50:34.610 Mustafa Raja: Does the, database agent handle universal inspectors?
480 00:50:37.230 ⇒ 00:50:39.289 Casie Aviles: Universal, oh, I think…
481 00:50:39.290 ⇒ 00:50:41.360 Amber Lin: We can add a column for that.
482 00:50:41.980 ⇒ 00:50:43.790 Casie Aviles: Yeah, we don’t, we don’t have,
483 00:50:44.220 ⇒ 00:50:47.259 Casie Aviles: Indication that something is universal or not.
484 00:50:47.840 ⇒ 00:50:50.359 Mustafa Raja: Okay, okay, can I share my screen for a moment?
485 00:50:51.500 ⇒ 00:50:52.879 Casie Aviles: Yeah, go ahead, go ahead.
486 00:50:54.940 ⇒ 00:51:01.939 Mustafa Raja: So right now, it sent back this. I think what it does is it, takes a look at, inspectors for this.
487 00:51:02.120 ⇒ 00:51:05.749 Mustafa Raja: Zip code, and then just mentions them, you know?
488 00:51:05.890 ⇒ 00:51:08.979 Mustafa Raja: So, we might also want to take care of that.
489 00:51:13.960 ⇒ 00:51:18.890 Casie Aviles: Okay, so what’s the expected, behavior.
490 00:51:18.890 ⇒ 00:51:26.029 Mustafa Raja: Yeah, so… that, I might not know. We might want to… This is a commercial instance.
491 00:51:26.030 ⇒ 00:51:27.110 Casie Aviles: Spectre.
492 00:51:28.400 ⇒ 00:51:33.260 Casie Aviles: Yeah, so, so, prior to this, yeah, it was sending this…
493 00:51:33.260 ⇒ 00:51:37.950 Mustafa Raja: So, so yeah, so the thing is, let me actually…
494 00:51:39.010 ⇒ 00:51:42.070 Mustafa Raja: Let me actually go in the execution.
495 00:53:02.920 ⇒ 00:53:07.889 Mustafa Raja: I’m using this test Andy, is it the same as this one?
496 00:53:19.210 ⇒ 00:53:22.180 Mustafa Raja: I saw that, Casey, you set up,
497 00:53:22.740 ⇒ 00:53:27.389 Mustafa Raja: a staging environment for Andy. Is this test Andy shifted to that?
498 00:53:36.330 ⇒ 00:53:37.599 Mustafa Raja: Okay, so don’t.
499 00:54:11.610 ⇒ 00:54:13.250 Mustafa Raja: Hey, Casey, didn’t you hear me?
500 00:54:13.780 ⇒ 00:54:14.350 Casie Aviles: Nope.
501 00:54:14.830 ⇒ 00:54:16.900 Casie Aviles: I think I got disconnected, sorry.
502 00:54:16.900 ⇒ 00:54:26.010 Mustafa Raja: Yeah, it’s okay. So, I was asking, is this ANDI test app connected to this, live brag agent?
503 00:54:26.660 ⇒ 00:54:29.950 Casie Aviles: Oh, no, no, it’s actually connected to the…
504 00:54:29.950 ⇒ 00:54:30.909 Mustafa Raja: The staging one, right?
505 00:54:30.910 ⇒ 00:54:31.719 Casie Aviles: Yeah, yeah.
506 00:54:32.440 ⇒ 00:54:35.740 Mustafa Raja: I think it’s not shared with me… could you share that…
507 00:54:36.350 ⇒ 00:54:37.670 Casie Aviles: Oh, sure, sure.
508 00:54:39.320 ⇒ 00:54:40.820 Casie Aviles: I’ll do that.
509 00:54:56.770 ⇒ 00:55:02.740 Casie Aviles: Okay. Yeah, I just shared it to you. I’m also gonna… Change the owner.
510 00:55:21.140 ⇒ 00:55:29.250 Mustafa Raja: Yeah, you see that, what this SQL does for Universal Inspector is it looks for all inspectors in the zip, you know?
511 00:55:29.900 ⇒ 00:55:35.470 Mustafa Raja: And it mentions all of them as, what’s it called? As Universal Inspectors.
512 00:55:35.840 ⇒ 00:55:46.390 Mustafa Raja: So, I think we don’t know universal inspectors for certain zips at all, so we might need to, you know.
513 00:55:46.640 ⇒ 00:55:50.260 Mustafa Raja: Get onboarded with them regarding the.
514 00:55:50.500 ⇒ 00:55:54.160 Casie Aviles: Okay, so basically we just want to clarify what universal…
515 00:55:54.500 ⇒ 00:55:56.469 Mustafa Raja: Yeah, yeah. And then so…
516 00:55:56.470 ⇒ 00:56:00.219 Casie Aviles: So, so that we can update R&D in our… okay, yeah.
517 00:56:00.550 ⇒ 00:56:01.370 Mustafa Raja: Yeah.
518 00:56:01.900 ⇒ 00:56:04.100 Mustafa Raja: Yeah, we just don’t have context on this.
519 00:56:06.270 ⇒ 00:56:16.859 Amber Lin: Yeah, we had a ticket a long time ago, we might as well just make a new one. It’s just a new column that we need to add and have them tag people.
520 00:56:20.640 ⇒ 00:56:22.970 Amber Lin: Cool, let’s see…
521 00:56:41.080 ⇒ 00:56:46.840 Amber Lin: Did Janice talk about universal inspectors, or should we ask her again to send that?
522 00:56:49.960 ⇒ 00:56:52.319 Casie Aviles: I, yeah, I remember there’s a ticket.
523 00:56:52.540 ⇒ 00:56:53.400 Casie Aviles: Somewhere.
524 00:56:59.450 ⇒ 00:57:02.459 Casie Aviles: Yeah, I think we can just ask again as well.
525 00:57:02.460 ⇒ 00:57:03.410 Mustafa Raja: Yeah.
526 00:57:03.570 ⇒ 00:57:06.570 Mustafa Raja: This triage is also from Janice.
527 00:57:19.160 ⇒ 00:57:23.470 Casie Aviles: Yeah, this is, I’m gonna s… Share.
528 00:57:23.730 ⇒ 00:57:25.360 Casie Aviles: This is.
529 00:57:26.330 ⇒ 00:57:36.540 Mustafa Raja: For sales lead… people, is the, zip code database agent responsible for that?
530 00:57:38.830 ⇒ 00:57:40.460 Casie Aviles: Oh… The biggest point.
531 00:57:41.040 ⇒ 00:57:42.680 Casie Aviles: If, if we…
532 00:57:42.680 ⇒ 00:57:44.520 Mustafa Raja: Let me share this screen again.
533 00:57:44.990 ⇒ 00:57:50.809 Casie Aviles: Yeah, so do you mean if we have, like, a classification, or of people who are sales leads?
534 00:57:51.630 ⇒ 00:58:01.320 Mustafa Raja: Yeah, yeah. So for this sort of question, who’s handling that? Is Andy handling that, or is, is the zip code agent handling that?
535 00:58:01.660 ⇒ 00:58:02.620 Mustafa Raja: This one.
536 00:58:04.030 ⇒ 00:58:07.789 Casie Aviles: I think… Yeah, I’m not.
537 00:58:07.790 ⇒ 00:58:13.700 Mustafa Raja: I feel… I feel this is more related to zip codes, you know, because… Corpus Christi.
538 00:58:13.700 ⇒ 00:58:14.140 Casie Aviles: Yeah.
539 00:58:14.140 ⇒ 00:58:21.340 Mustafa Raja: And I think, let me actually see where… how it got routed.
540 00:58:23.460 ⇒ 00:58:26.109 Mustafa Raja: Did it call the Zips TV? I think… Yeah.
541 00:58:26.110 ⇒ 00:58:32.010 Casie Aviles: If we, if we tell, you know, the agent, the query agent to
542 00:58:32.940 ⇒ 00:58:39.670 Casie Aviles: consider sales lead as just basically as inspector, I mean, then that.
543 00:58:39.670 ⇒ 00:58:45.730 Mustafa Raja: Yeah, that should be good, yeah. Yeah, right now, it’s just putting in, looking for role name sales.
544 00:58:46.090 ⇒ 00:58:46.829 Mustafa Raja: Yeah, I think…
545 00:58:46.830 ⇒ 00:58:47.500 Casie Aviles: We don’t have that.
546 00:58:47.500 ⇒ 00:58:52.999 Mustafa Raja: Minor… Minor prompt is… what’s it called? Prompt adjustment?
547 00:58:53.240 ⇒ 00:58:55.229 Mustafa Raja: Yeah.
548 00:58:55.230 ⇒ 00:58:55.650 Casie Aviles: So you…
549 00:58:55.650 ⇒ 00:58:56.429 Mustafa Raja: We can go to…
550 00:58:56.430 ⇒ 00:58:57.640 Casie Aviles: Long fees to do that.
551 00:58:58.220 ⇒ 00:59:04.850 Mustafa Raja: Yeah, so, for sales lead, those are… those are actually inspectors.
552 00:59:06.510 ⇒ 00:59:09.920 Casie Aviles: Yeah, I believe so, that that should be inspectors.
553 00:59:09.920 ⇒ 00:59:12.969 Amber Lin: Yeah, sales and inspectors are the same thing.
554 00:59:15.910 ⇒ 00:59:17.430 Mustafa Raja: I think I’m having some.
555 00:59:17.430 ⇒ 00:59:28.949 Amber Lin: Sorry, sorry. Sales lead… the lead is their customer. A lead is, like, a lead that… that has interest in our services, right? Sales.
556 00:59:29.060 ⇒ 00:59:38.389 Amber Lin: the inspectors do sales, so if they ask who can do sales in this area, they’re asking for an inspector. But a lead is just, like.
557 00:59:38.720 ⇒ 00:59:39.959 Amber Lin: A customer. A customer.
558 00:59:39.960 ⇒ 00:59:40.690 Mustafa Raja: Yeah.
559 00:59:40.690 ⇒ 00:59:47.179 Amber Lin: But if they ask… if I have a commercial sales lead, it’s like, hey, I have this commercial.
560 00:59:48.220 ⇒ 00:59:48.740 Mustafa Raja: Justin…
561 00:59:48.740 ⇒ 00:59:55.410 Amber Lin: That I… that I need an inspector for, so yes, they are, like, who should… they ask, like, who should I put it on? They’re asking for an inspector.
562 00:59:55.540 ⇒ 00:59:57.520 Amber Lin: Sorry, it was complicated. They’re asking for.
563 00:59:57.520 ⇒ 00:59:58.350 Mustafa Raja: Yeah.
564 00:59:59.360 ⇒ 01:00:00.929 Casie Aviles: Yeah, that makes sense now.
565 01:00:02.090 ⇒ 01:00:02.740 Casie Aviles: So…
566 01:00:02.740 ⇒ 01:00:04.409 Mustafa Raja: I’ll update this, I’ll update this.
567 01:00:05.220 ⇒ 01:00:15.500 Casie Aviles: Yeah, basically you just have to add there that… and, it should be… when it’s asking for commercial sales lead, then we should look for inspectors.
568 01:00:16.270 ⇒ 01:00:17.060 Mustafa Raja: Yeah.
569 01:00:18.180 ⇒ 01:00:18.920 Casie Aviles: Okay.
570 01:00:23.340 ⇒ 01:00:25.170 Casie Aviles: Yeah, that should be a long fuse fix.
571 01:00:26.180 ⇒ 01:00:30.920 Mustafa Raja: Yeah, I’ll update that, I just have some sorts of connectivity issue with languages right now.
572 01:00:32.740 ⇒ 01:00:33.340 Casie Aviles: Okay.
573 01:01:30.780 ⇒ 01:01:35.580 Amber Lin: Cool, okay. If you guys have, like, extra time, we can…
574 01:01:35.710 ⇒ 01:01:40.169 Amber Lin: keep working on the triage tickets, but I have to hop right now.
575 01:01:40.800 ⇒ 01:01:41.510 Mustafa Raja: Okay.
576 01:01:41.840 ⇒ 01:01:42.450 Amber Lin: Cool.
577 01:01:43.110 ⇒ 01:01:48.599 Amber Lin: Thank you both. I asked Tara about the dispatch stuff, so Casey, I’ll get back to you on that.
578 01:01:49.840 ⇒ 01:01:50.940 Casie Aviles: Alright, thank you.
579 01:01:51.580 ⇒ 01:01:52.420 Amber Lin: Bye.