Meeting Title: Brainforge Element EP Work Sync Date: 2026-03-03 Meeting participants: Brylle Girang, Amber Lin
WEBVTT
1 00:02:59.310 ⇒ 00:03:00.410 Amber Lin: Hi there!
2 00:03:00.620 ⇒ 00:03:01.870 Brylle Girang: Hi, Amber!
3 00:03:02.710 ⇒ 00:03:04.180 Amber Lin: Hi, how are you?
4 00:03:04.510 ⇒ 00:03:09.499 Brylle Girang: I am doing great. It’s Wednesday here, and it is so packed.
5 00:03:10.070 ⇒ 00:03:12.089 Brylle Girang: Oh, it’s Wednesday! Wow.
6 00:03:12.920 ⇒ 00:03:16.170 Amber Lin: Yeah, I mean, it’s Tuesday for me.
7 00:03:16.320 ⇒ 00:03:17.270 Brylle Girang: So…
8 00:03:17.430 ⇒ 00:03:24.680 Brylle Girang: Oh, it’s the way up here. Okay, I was confused, I thought that the days just went by. How are you doing?
9 00:03:24.680 ⇒ 00:03:25.340 Amber Lin: Only.
10 00:03:26.190 ⇒ 00:03:27.819 Amber Lin: I’m good.
11 00:03:27.990 ⇒ 00:03:31.429 Amber Lin: I worked out yesterday, so I’m really, really, really sore.
12 00:03:31.900 ⇒ 00:03:34.440 Brylle Girang: Work out like, you went to the gym?
13 00:03:34.740 ⇒ 00:03:38.890 Amber Lin: Yeah, I lifted yesterday, so I’m really sore.
14 00:03:39.140 ⇒ 00:03:40.550 Brylle Girang: When did you start?
15 00:03:41.120 ⇒ 00:03:48.190 Amber Lin: I started going to the gym, I think, like, Two, three years ago?
16 00:03:48.190 ⇒ 00:03:49.529 Brylle Girang: Oh, so you…
17 00:03:49.530 ⇒ 00:03:51.649 Amber Lin: Oh, Okay. Yeah.
18 00:03:52.330 ⇒ 00:03:54.820 Brylle Girang: We just started going to the gym, like.
19 00:03:56.160 ⇒ 00:03:56.640 Amber Lin: Whoa.
20 00:03:56.640 ⇒ 00:03:58.670 Brylle Girang: October, and then…
21 00:03:58.670 ⇒ 00:03:59.530 Amber Lin: Whoa!
22 00:03:59.530 ⇒ 00:04:02.740 Brylle Girang: I just started… Doing calorie deficit?
23 00:04:03.000 ⇒ 00:04:05.260 Brylle Girang: last January, so from.
24 00:04:05.260 ⇒ 00:04:07.649 Amber Lin: I’m… I’m so impressed.
25 00:04:07.650 ⇒ 00:04:10.249 Brylle Girang: It’s so hard to do a deficit.
26 00:04:10.380 ⇒ 00:04:18.829 Brylle Girang: It is, and that’s why from October to December, nothing really happened. It was just 3 months worth of sore muscles, and then I didn’t lose
27 00:04:20.240 ⇒ 00:04:27.729 Brylle Girang: But… it feels energizing, it feels like I have life outside of, you know, sitting in the office.
28 00:04:27.730 ⇒ 00:04:34.210 Amber Lin: I know, I know, I’m either at home, or I’m at the gym, and I can’t find… because I’m trying to get…
29 00:04:34.740 ⇒ 00:04:38.979 Amber Lin: out of using Instagram, because I scroll too much.
30 00:04:39.650 ⇒ 00:04:49.529 Amber Lin: But I don’t scroll, I have nothing to do, so I’m either working, or I’m just at the gym, and then I come home and I don’t know what to do.
31 00:04:50.150 ⇒ 00:04:58.980 Brylle Girang: I removed all my social media one time, just because of that same problem. I was doom scrolling for, like, 1 hour, 2 hours, and then…
32 00:04:58.980 ⇒ 00:05:00.110 Amber Lin: Mmm…
33 00:05:00.630 ⇒ 00:05:02.370 Brylle Girang: My brain was rotting.
34 00:05:02.470 ⇒ 00:05:09.560 Brylle Girang: Anyways, yeah, so, we have two main topics today. The first one is the…
35 00:05:09.970 ⇒ 00:05:20.160 Brylle Girang: EP work for Element, and I also had this conversation with Awash, so right now, I’m just, you know, sitting in as a temporary EP, just trying to
36 00:05:20.810 ⇒ 00:05:24.990 Brylle Girang: understand what you guys are doing, just trying to understand where I can help.
37 00:05:25.110 ⇒ 00:05:25.710 Brylle Girang: But…
38 00:05:26.330 ⇒ 00:05:43.789 Brylle Girang: I’m… I do think that within a month or so, I will be replaced by a new hire. I’m not really sure, that’s what Utam has mentioned. But in the meantime, while I am sitting as an EP, I think the most viable
39 00:05:43.840 ⇒ 00:05:50.309 Brylle Girang: Distinction between our roles is that you will handle anything related to analytics.
40 00:05:50.400 ⇒ 00:05:55.430 Brylle Girang: And then I will handle the EP work related to data engineering.
41 00:05:56.270 ⇒ 00:05:58.430 Amber Lin: Yeah, sounds good.
42 00:05:58.620 ⇒ 00:06:10.410 Amber Lin: So, I guess the only clarification is, if any of my analytics work needs modeling, which I would say is more, like, AE work.
43 00:06:10.410 ⇒ 00:06:12.180 Brylle Girang: That would so be me, right?
44 00:06:12.360 ⇒ 00:06:15.670 Amber Lin: Because it’s from my analytics reporting needs.
45 00:06:16.260 ⇒ 00:06:23.769 Brylle Girang: I think it’s going to be under mine, and I’m just going… I’m going to create, like, a linear workflow.
46 00:06:24.110 ⇒ 00:06:37.879 Brylle Girang: Ideally, you can just, you know, trigger a command, or, like, call linear, and then create a modeling-specific ticket that will be assigned to Awash, and then I should be the one pushing for that.
47 00:06:38.280 ⇒ 00:06:39.460 Amber Lin: Okay, yeah.
48 00:06:39.460 ⇒ 00:06:42.170 Brylle Girang: as possible, I want our tickets to…
49 00:06:42.450 ⇒ 00:06:52.169 Brylle Girang: be separated, because in the future, once we do some actual analytics on our… on our tasks, it will be hard to gauge how…
50 00:06:52.170 ⇒ 00:06:55.679 Amber Lin: I see. Okay. That works for me.
51 00:06:56.170 ⇒ 00:07:01.230 Amber Lin: Do you want to just send a quick message about that in a channel? Yep.
52 00:07:01.700 ⇒ 00:07:12.319 Brylle Girang: Okay, perfect. And then, okay, I think that’s clear, and then the second part here is, like, understanding how we can use cursor for your EP work, right?
53 00:07:12.630 ⇒ 00:07:20.360 Amber Lin: Yeah, I don’t think we need a full hour, but I just want to know what’s possible, because I don’t think I’ve even thought about some of the things before.
54 00:07:20.360 ⇒ 00:07:21.930 Brylle Girang: Yeah, it’s okay. I,
55 00:07:22.200 ⇒ 00:07:28.949 Brylle Girang: maybe we can start with you just walking me through how you use Cursor, or what your…
56 00:07:29.180 ⇒ 00:07:31.179 Brylle Girang: day-to-day tasks are. Is that something.
57 00:07:31.180 ⇒ 00:07:43.730 Amber Lin: Yeah, I think we talked a little bit about this last time. I don’t really use it much for anything outside my analytics work, and for my analytics work, I connected to BigQuery, and I…
58 00:07:43.730 ⇒ 00:07:56.290 Amber Lin: instructor to do the analysis. The only thing, really, outside of analytics that I’ve used it for was for, say, writing a analytics plan, writing a proposal to send out.
59 00:07:57.380 ⇒ 00:08:06.950 Amber Lin: So, I think there’s a lot more, I think there’s a lot on the platform, so if you can maybe share a screen and walk me through different things, that would be great.
60 00:08:07.320 ⇒ 00:08:15.290 Brylle Girang: Okay, and can you confirm that your cursor has been set up properly? Like, you can access the fold, you can access the platform, etc?
61 00:08:15.290 ⇒ 00:08:21.800 Amber Lin: Yeah, I think so. One more thing, I think my cursor’s on my, like, personal cursor account.
62 00:08:22.310 ⇒ 00:08:30.790 Amber Lin: Like, I had Premium Cursor before, I don’t think I ever canceled it. Are we using, like, a company workspace? Is that important?
63 00:08:30.790 ⇒ 00:08:32.080 Brylle Girang: Yeah, and actually.
64 00:08:32.080 ⇒ 00:08:32.780 Amber Lin: Yeah, I think…
65 00:08:32.780 ⇒ 00:08:39.040 Brylle Girang: the main reason why your usage… your usage rate when I did the report was zero.
66 00:08:39.340 ⇒ 00:08:40.590 Brylle Girang: Because you were using your.
67 00:08:40.590 ⇒ 00:08:43.159 Amber Lin: Oh, I see, I see, okay.
68 00:08:43.169 ⇒ 00:08:46.809 Brylle Girang: Let me ask Rico to re-invite you over.
69 00:08:47.040 ⇒ 00:08:48.539 Amber Lin: Okay, that would be great.
70 00:08:49.410 ⇒ 00:08:55.110 Brylle Girang: Yeah, okay, so we… oh, sorry. We, basically have the…
71 00:08:55.940 ⇒ 00:09:04.700 Brylle Girang: very cool here, right? The Brainforge platform, and this is where we store everything related to our clients, and I think the most helpful.
72 00:09:04.700 ⇒ 00:09:05.670 Amber Lin: is…
73 00:09:05.880 ⇒ 00:09:18.060 Brylle Girang: the meeting transcripts. But I’m also working with OTAM. I am personally leading the charge into consolidating our tools, and that involves, you know, Instagant.
74 00:09:18.290 ⇒ 00:09:22.520 Brylle Girang: operating, and I’m pushing for…
75 00:09:22.810 ⇒ 00:09:30.839 Brylle Girang: us to either just use linear for everything, or maybe we can switch to Asana, which will be much more helpful for us, right?
76 00:09:31.490 ⇒ 00:09:34.049 Brylle Girang: Because the main problem here is that
77 00:09:34.240 ⇒ 00:09:38.300 Brylle Girang: Instagant and operating… well, Instagant is not connected to
78 00:09:39.180 ⇒ 00:09:52.430 Brylle Girang: cursor, and there’s no way right now to connect Instagant to cursor, and that might be helpful, like, gauging timelines. Here are linear tasks, how are we right now with our timelines, those kinds of questions.
79 00:09:53.720 ⇒ 00:10:00.800 Brylle Girang: We’re a bit limited on that, but we’re moving aggressively into, like, fixing that within the next 2 to 3 months. But…
80 00:10:00.800 ⇒ 00:10:01.560 Amber Lin: I see.
81 00:10:02.300 ⇒ 00:10:05.049 Brylle Girang: At the same time, I, I think…
82 00:10:05.390 ⇒ 00:10:11.720 Brylle Girang: And I believe that, I don’t know if it’s you or Mustafa who mentioned that it’s a bit hard to
83 00:10:12.080 ⇒ 00:10:27.610 Brylle Girang: using Notion, or managing Notion right now, because we have the vote, we have Notion, we have the meeting transcripts, and then we need to do everything in Google Docs, etc, etc. So Notion is another friction point that I’m trying to
84 00:10:27.740 ⇒ 00:10:30.530 Brylle Girang: like, I’m trying to help out
85 00:10:31.050 ⇒ 00:10:34.950 Brylle Girang: consolidate into one tool. So it’s either no…
86 00:10:34.950 ⇒ 00:10:35.560 Amber Lin: tested?
87 00:10:36.030 ⇒ 00:10:40.280 Brylle Girang: or we maybe moved to Obsidian. Have you… have you heard of Obsidian?
88 00:10:40.280 ⇒ 00:10:57.370 Amber Lin: I really like Obsidian, I use it personally, but I do think it’s not more of a social sharing tool. It’s harder to have, like, images… it’s more code-like, and I just don’t know how easy it is to share stuff with clients.
89 00:10:57.980 ⇒ 00:11:02.889 Brylle Girang: Yeah, well, when it comes to sharing stuff with clients, that’s actually one thing that I’m…
90 00:11:03.390 ⇒ 00:11:16.619 Brylle Girang: trying to understand, because if that’s going to be our main reason for sticking with Notion, I would say we have Google Docs, and most of the things that we’re really sharing with our clients are in Google Docs.
91 00:11:16.730 ⇒ 00:11:22.359 Brylle Girang: I think just a little percentage of those are in Notion, but yeah, so…
92 00:11:22.360 ⇒ 00:11:22.980 Amber Lin: So…
93 00:11:23.430 ⇒ 00:11:27.060 Brylle Girang: We’re… we’re checking, or I’m trying to see if, you know, we can move
94 00:11:27.240 ⇒ 00:11:36.870 Brylle Girang: to Obsygen instead. So Obsygen works basically like a GitHub repo, right? We have a folder, and that’s where everything lives. So that’s going to be.
95 00:11:36.870 ⇒ 00:11:37.250 Amber Lin: Yeah.
96 00:11:37.250 ⇒ 00:11:46.869 Brylle Girang: easy to sync. But yeah, just sharing with you what our future plans are, which I’m pretty sure will also help us with the challenges that we have mentioned.
97 00:11:47.540 ⇒ 00:11:54.060 Brylle Girang: But in the meantime, here’s where we currently are at. We have two main commands.
98 00:11:54.180 ⇒ 00:11:58.819 Brylle Girang: That… that has been established, like, last week, and within this week.
99 00:11:58.940 ⇒ 00:12:08.820 Brylle Girang: which focuses on EP-related work. I think the main… one of the main commands here would be the EP audit command.
100 00:12:10.140 ⇒ 00:12:17.889 Brylle Girang: So, you can read… you can read this SOP, and basically what it just does is it makes sure that everything has an
101 00:12:18.260 ⇒ 00:12:19.929 Brylle Girang: Everything has a linear ticket.
102 00:12:20.220 ⇒ 00:12:29.680 Brylle Girang: So, after you’re done with the meeting, after maybe Robert or Element is done with the meeting, it gets pulled through our bulk vault.
103 00:12:29.890 ⇒ 00:12:36.629 Brylle Girang: And this command makes sure that it studies those transcripts, and then cross-checks it with linear.
104 00:12:37.710 ⇒ 00:12:46.620 Brylle Girang: The main goal of this EP audit command is to make sure that we are on track when it comes to our linear boards. There’s nothing missing.
105 00:12:46.770 ⇒ 00:12:49.369 Brylle Girang: Everything’s documented via a ticket.
106 00:12:49.570 ⇒ 00:12:56.449 Brylle Girang: And if possible, existing tickets will be updated based on the updates from the transcripts.
107 00:12:57.070 ⇒ 00:13:05.820 Brylle Girang: So if you… if you run this without anything, don’t worry, because we… we configured this command, to be…
108 00:13:06.570 ⇒ 00:13:13.529 Brylle Girang: to be guided. Like, if you just run it, it will ask you to, like, give us, or give, give, give, give…
109 00:13:13.590 ⇒ 00:13:14.990 Amber Lin: the AI.
110 00:13:14.990 ⇒ 00:13:15.730 Brylle Girang: the specific.
111 00:13:15.730 ⇒ 00:13:16.460 Amber Lin: decline.
112 00:13:16.460 ⇒ 00:13:17.630 Brylle Girang: You want to target.
113 00:13:17.740 ⇒ 00:13:21.329 Brylle Girang: So just for the sake of this, I think I’m going to go with Element.
114 00:13:24.700 ⇒ 00:13:30.260 Brylle Girang: And then what it does is, this is divided into basically 9 steps.
115 00:13:30.380 ⇒ 00:13:36.780 Brylle Girang: The first step is it searches the vault for everything related to Element.
116 00:13:36.910 ⇒ 00:13:43.090 Brylle Girang: transcripts from clients, granola summaries. If possible.
117 00:13:43.340 ⇒ 00:13:56.040 Brylle Girang: new files or new documents related to Element. And at the same time, it also searches our standups, where the client’s name is rarely mentioned.
118 00:13:56.220 ⇒ 00:14:00.300 Brylle Girang: And it tries to see which of those relates to element.
119 00:14:01.070 ⇒ 00:14:02.170 Amber Lin: I see.
120 00:14:02.170 ⇒ 00:14:20.100 Brylle Girang: That one is a pre-audit. You can see it here. This is just going to be the AI asking us, is everything covered? Because there might be instances that there’s a missing meeting, or there’s something that has been discussed of the meetings that you want to add.
121 00:14:20.330 ⇒ 00:14:26.599 Brylle Girang: But this is just a quick check, making sure that before it proceeds with the audit, everything has been covered.
122 00:14:26.900 ⇒ 00:14:29.340 Brylle Girang: At the same time, it gives you a quick summary.
123 00:14:29.500 ⇒ 00:14:32.270 Brylle Girang: If you want to get up to speed.
124 00:14:32.530 ⇒ 00:14:38.749 Brylle Girang: Right? So, this checks linear, this checks the transcripts, this checks the voltage.
125 00:14:38.750 ⇒ 00:14:40.890 Amber Lin: Does this check Slack at all?
126 00:14:41.310 ⇒ 00:14:53.420 Brylle Girang: That is something that I’m working with Utam. I’m moving aggressively there, because I understand you. There are so many things going on in Slack that are not documented, right?
127 00:14:53.640 ⇒ 00:15:08.219 Brylle Girang: My temporary workaround for that is I’m just going to the Slack channels, I’m creating the summary, and then I’m pasting it here, just so that it’s able to also gather that information.
128 00:15:08.590 ⇒ 00:15:09.270 Brylle Girang: But that’s…
129 00:15:09.270 ⇒ 00:15:12.150 Amber Lin: What do you mean by creating a summary in Slack?
130 00:15:12.150 ⇒ 00:15:19.160 Brylle Girang: Yeah, so let me… Just a moment. Let me try and share my screen.
131 00:15:19.530 ⇒ 00:15:23.689 Amber Lin: Is that the Slack client hub agents, or…
132 00:15:24.610 ⇒ 00:15:25.290 Brylle Girang: No.
133 00:15:25.290 ⇒ 00:15:27.520 Amber Lin: using ropes? Oh, okay.
134 00:15:29.030 ⇒ 00:15:36.890 Brylle Girang: Just a moment, yeah, so let’s say… Oh.
135 00:15:38.770 ⇒ 00:15:41.669 Brylle Girang: I’m in the client element channel, and then…
136 00:15:42.440 ⇒ 00:15:46.710 Brylle Girang: If you just click the ellipses, and then click Summarize.
137 00:15:47.180 ⇒ 00:15:53.239 Amber Lin: Oh, wow. Okay, I’ve never tried that. I’m so behind. That’s very cool.
138 00:15:53.650 ⇒ 00:16:04.539 Amber Lin: Yeah, and it doesn’t… it’s safer, because it doesn’t give the summary over to the clients, right? Or… it’s different from the bots, where it will just dump the summary over to the… to the messages.
139 00:16:04.740 ⇒ 00:16:09.910 Brylle Girang: So I’m just making sure that I click more details so everything is covered.
140 00:16:11.370 ⇒ 00:16:12.210 Brylle Girang: Just so you know…
141 00:16:12.210 ⇒ 00:16:12.590 Amber Lin: Awesome.
142 00:16:12.590 ⇒ 00:16:13.600 Brylle Girang: space this.
143 00:16:14.260 ⇒ 00:16:15.629 Amber Lin: Okay.
144 00:16:16.250 ⇒ 00:16:17.740 Brylle Girang: Paste this over to cursor.
145 00:16:17.940 ⇒ 00:16:26.209 Brylle Girang: So, yeah. So that’s what I’m doing, personally. I’m not going to do that just for the sake of my context window here, but…
146 00:16:26.500 ⇒ 00:16:37.950 Brylle Girang: Yeah, so it’s a pre-audit first, just trying to confirm if everything’s good. There’s a confirmation message here, and if you want to add something, you know, if you want to…
147 00:16:38.290 ⇒ 00:16:44.069 Brylle Girang: maybe tell AI that there’s something else that’s missing, etc, then please do so.
148 00:16:44.320 ⇒ 00:16:48.179 Brylle Girang: Just going to tell… A cursor to proceed.
149 00:16:54.450 ⇒ 00:17:00.649 Brylle Girang: And then it will run the full audit. So this is also divided into steps. The first step is…
150 00:17:00.790 ⇒ 00:17:09.459 Brylle Girang: That it checks linear for existing tickets, and updates them if there are updates mentioned in the meetings or in the chats.
151 00:17:10.089 ⇒ 00:17:15.329 Amber Lin: All of them need to be in the linear project, right?
152 00:17:15.690 ⇒ 00:17:16.470 Brylle Girang: Sorry.
153 00:17:16.910 ⇒ 00:17:24.009 Amber Lin: What does it check for in linear? So, everything under the element team?
154 00:17:24.339 ⇒ 00:17:28.009 Brylle Girang: Yep, so everything under the element team, you can be…
155 00:17:28.249 ⇒ 00:17:36.919 Brylle Girang: you can, you can drill down, you can, you can tell cursor to, hey, just look at the data analytics projects, etc. So…
156 00:17:38.010 ⇒ 00:17:42.420 Amber Lin: But it will only look at the client projects that you mentioned.
157 00:17:42.580 ⇒ 00:17:43.080 Brylle Girang: From the.
158 00:17:43.080 ⇒ 00:17:43.930 Amber Lin: Hmm, okay.
159 00:17:44.540 ⇒ 00:17:51.009 Brylle Girang: So, as you can see here, it also gives some thought process. So it searches the wholesale renewal issues, then…
160 00:17:51.670 ⇒ 00:17:53.159 Amber Lin: Hmm, okay.
161 00:17:53.160 ⇒ 00:18:02.279 Brylle Girang: One thing, it updates the descriptions, or adds… adds comment… add comments, if there… if there are updates. The second step is that
162 00:18:02.530 ⇒ 00:18:07.519 Brylle Girang: It assigns the tickets, just in case there are no assignees yet.
163 00:18:07.700 ⇒ 00:18:13.250 Brylle Girang: And it does it intelligently, in a way that it checks the related tickets, it checks the transcripts.
164 00:18:13.250 ⇒ 00:18:13.810 Amber Lin: Hmm.
165 00:18:13.810 ⇒ 00:18:17.700 Brylle Girang: And then whoever’s the closest, it will assign it to that person.
166 00:18:17.700 ⇒ 00:18:24.580 Amber Lin: I see. Is there a way for me to check the assignments? I want to make sure that it’s correct. Sometimes it’ll do it wrong.
167 00:18:24.580 ⇒ 00:18:29.949 Brylle Girang: Yeah, yeah, after the whole steps are done, it will give us a summary of whatever.
168 00:18:29.950 ⇒ 00:18:31.490 Amber Lin: Okay, okay, awesome.
169 00:18:31.680 ⇒ 00:18:32.490 Brylle Girang: So…
170 00:18:32.790 ⇒ 00:18:38.480 Brylle Girang: You can… you can check linear, or you can check the summary, and then you can just update the linear ticket if it’s wrong.
171 00:18:40.210 ⇒ 00:18:50.869 Brylle Girang: And then lastly, I think one of the most helpful for me is that it just gives a summary of what has happened, and then this is something that, you know, you can paste over to…
172 00:18:51.690 ⇒ 00:18:56.680 Brylle Girang: To the Slack channels, to the client channels, or to our internal Slack channels.
173 00:18:58.210 ⇒ 00:19:00.869 Brylle Girang: So for element, here are the stuff.
174 00:19:01.490 ⇒ 00:19:08.409 Brylle Girang: So, everything that is in here are the stuff that we have mentioned during the meetings.
175 00:19:09.050 ⇒ 00:19:09.660 Amber Lin: Seriously.
176 00:19:09.660 ⇒ 00:19:11.410 Brylle Girang: there’s wholesale.
177 00:19:11.550 ⇒ 00:19:18.579 Brylle Girang: There’s the e-commerce modeling for Awash, and it also tells us if there are tickets that have been
178 00:19:18.770 ⇒ 00:19:22.129 Brylle Girang: That have been created, or tickets that have been updated.
179 00:19:22.360 ⇒ 00:19:25.290 Brylle Girang: But mostly it’s just, you know, keeping us in track.
180 00:19:25.400 ⇒ 00:19:27.030 Brylle Girang: Of the current tickets.
181 00:19:27.390 ⇒ 00:19:28.970 Amber Lin: I see, okay.
182 00:19:30.680 ⇒ 00:19:44.529 Brylle Girang: And at the same time, yeah, there… it also highlights the risks and escalations. For example, there’s a ticket that has been mentioned during the meetings. It will tell you, or if there are tickets that are already overdue, or still blocked.
183 00:19:44.530 ⇒ 00:19:46.570 Amber Lin: Oh, awesome, okay.
184 00:19:47.950 ⇒ 00:19:58.289 Brylle Girang: to here, you can see in step 3, it tells us if there have been created issues. It tells us that there’s none, because all the action items have been.
185 00:19:58.290 ⇒ 00:19:59.199 Amber Lin: Call me.
186 00:19:59.360 ⇒ 00:20:01.270 Brylle Girang: With the existing tickets, but…
187 00:20:01.270 ⇒ 00:20:01.850 Amber Lin: Okay.
188 00:20:01.850 ⇒ 00:20:07.290 Brylle Girang: We have added a ticket for element 107, based on the kickoff call.
189 00:20:08.030 ⇒ 00:20:10.870 Amber Lin: Okay. Awesome.
190 00:20:12.090 ⇒ 00:20:20.769 Amber Lin: Another question before we proceed to the other command. So, I know our platform gets updated every so often.
191 00:20:20.770 ⇒ 00:20:32.480 Amber Lin: When I… say, when I do the poll again, so I pull the most updated version, does it overwrite all the documents I’ve created so far?
192 00:20:33.480 ⇒ 00:20:36.510 Brylle Girang: By document… what do you mean by documents?
193 00:20:36.980 ⇒ 00:20:50.849 Amber Lin: So I… this is less related to EP tasks, but say if it was in the chat, and I created, like, a markdown or an analysis file that I wanted to send out, but I need to pull the…
194 00:20:51.380 ⇒ 00:21:00.419 Amber Lin: main branch again, does it overwrite my stuff, or do I have to, like, submit… commit a PR before I do that, and how does that work?
195 00:21:00.600 ⇒ 00:21:05.899 Brylle Girang: No, no, as long as it is a new file, and it’s not, you know.
196 00:21:06.160 ⇒ 00:21:21.770 Brylle Girang: manipulated by the new poll, then it won’t be overwritten. So if it’s just, you know, a file that’s on your local drive right now, that you haven’t committed yet, that you haven’t pushed yet, don’t worry, you can just pull the latest updates, and it will not be affected.
197 00:21:22.200 ⇒ 00:21:33.220 Amber Lin: Okay, awesome, because I was worried about that. I’m using the earlier platform version. And confirming, are we using Blainforge platform or BrainForge Data Platform?
198 00:21:33.220 ⇒ 00:21:39.640 Brylle Girang: Yeah, BrainForge platform, so this is what we’re using. I just have this for the codes, but it’s not.
199 00:21:39.640 ⇒ 00:21:40.100 Amber Lin: Hmm.
200 00:21:40.100 ⇒ 00:21:42.190 Brylle Girang: It’s not really the main repo.
201 00:21:42.190 ⇒ 00:21:43.910 Amber Lin: Okay, awesome.
202 00:21:43.910 ⇒ 00:21:44.860 Brylle Girang: I’ll keep mine.
203 00:21:45.260 ⇒ 00:21:56.599 Amber Lin: Sounds good. Anything I should look at within the Brainforge platform? Should I take a look at the playbook? Take a look at the fall? Where… where are the client,
204 00:21:56.710 ⇒ 00:21:58.070 Amber Lin: folders.
205 00:21:58.070 ⇒ 00:22:06.329 Brylle Girang: Yeah, no, you can take a look, but you can just ask cursor, right, where, what’s in the element folder.
206 00:22:07.990 ⇒ 00:22:09.030 Amber Lin: I see.
207 00:22:09.160 ⇒ 00:22:17.569 Brylle Girang: And this is also, you know, one of the main things why I’m pushing for Obsidian. If you’re searching for something, you can just ask Cursor, and then it will take you over to the.
208 00:22:17.570 ⇒ 00:22:22.649 Amber Lin: I see, awesome. Quick question for this, I know you clicked ask.
209 00:22:24.120 ⇒ 00:22:31.779 Amber Lin: is… how do you select between the different tools? There’s ASK, and I don’t know there’s something else, like, how would you…
210 00:22:32.130 ⇒ 00:22:34.479 Brylle Girang: Yeah. How do you select between them?
211 00:22:34.770 ⇒ 00:22:37.049 Brylle Girang: I’m just going to,
212 00:22:38.410 ⇒ 00:22:48.030 Brylle Girang: use some… how do you call this? Some metaphors here. So, as is just an AI with, like, a mouth and a brain.
213 00:22:48.130 ⇒ 00:22:51.300 Brylle Girang: It has no hands, so it can’t do anything.
214 00:22:51.820 ⇒ 00:22:52.580 Amber Lin: Bye.
215 00:22:52.580 ⇒ 00:22:54.380 Brylle Girang: I, I just subconsciously
216 00:22:54.640 ⇒ 00:23:01.270 Brylle Girang: I chose as, because I was not expecting the AI to, like, produce an output for me.
217 00:23:02.660 ⇒ 00:23:15.519 Brylle Girang: If you want… if you want to make sure that the AI does not do anything without your consent, use us. It may be just general questions, or, like, brainstorming, just brainstorming with cursors.
218 00:23:15.520 ⇒ 00:23:18.060 Amber Lin: Oh, I see. That’s helpful.
219 00:23:18.060 ⇒ 00:23:19.629 Brylle Girang: Something, etc, right?
220 00:23:19.740 ⇒ 00:23:23.739 Brylle Girang: So, when you choose agent, you’re giving the AI
221 00:23:23.950 ⇒ 00:23:26.919 Brylle Girang: Like, to actually produce an output.
222 00:23:27.910 ⇒ 00:23:33.050 Brylle Girang: Like, hey, I need help, but I want you to update the files, etc.
223 00:23:33.730 ⇒ 00:23:37.889 Brylle Girang: So it can do that, so it can actually do things for us.
224 00:23:39.740 ⇒ 00:23:47.450 Brylle Girang: And then plan is, like, giving your AI with an additional set of brains. Like, it just makes it more intelligent.
225 00:23:47.620 ⇒ 00:23:48.240 Brylle Girang: So…
226 00:23:48.240 ⇒ 00:23:52.950 Amber Lin: How is it different from the regular agent? So is this just better at planning?
227 00:23:53.300 ⇒ 00:23:58.820 Brylle Girang: Yeah, so, there is this thing called prompt engineering. I’m pretty sure that you have heard of that.
228 00:23:59.950 ⇒ 00:24:08.379 Brylle Girang: So, the way that these LLMs work is that the more detailed the instructions are, the more
229 00:24:09.290 ⇒ 00:24:11.759 Brylle Girang: Relevant their outputs will be.
230 00:24:12.830 ⇒ 00:24:15.070 Brylle Girang: So, plan helps us create.
231 00:24:15.350 ⇒ 00:24:18.430 Brylle Girang: a more detailed prompt for the AI.
232 00:24:18.680 ⇒ 00:24:19.330 Brylle Girang: So…
233 00:24:19.330 ⇒ 00:24:36.669 Amber Lin: I see, I see. I think my problem I came across when I was using Plan was that because it planned so far ahead, when it actually was executing, some of the steps weren’t perfect, but because
234 00:24:36.670 ⇒ 00:24:40.200 Amber Lin: AI has already structured the whole plan,
235 00:24:40.750 ⇒ 00:24:52.339 Amber Lin: it actually got worse results than just doing it step by step by step and refining each step before proceeding. So that was what I came across from using the plan, so…
236 00:24:52.350 ⇒ 00:25:01.479 Amber Lin: like, I don’t know if I’m using it for the right thing, or is there other ways you recommend using the plan feature? Because I do think it’s quite nice.
237 00:25:01.840 ⇒ 00:25:17.769 Brylle Girang: Yeah, I do think that iterating is a huge part of this, so if we just give, you know, cursor, like, a two-sentence prompt for it to create a plan, I’m pretty sure that it will not be as perfect as we want it to be.
238 00:25:18.120 ⇒ 00:25:31.260 Brylle Girang: So if you… if we won’t be able to give them, like, the whole context, what we want to happen, then at least we need to iterate whatever it provides us. Let’s say… let’s try it out. So I need…
239 00:25:31.650 ⇒ 00:25:36.510 Brylle Girang: hmm, I need to create…
240 00:25:36.760 ⇒ 00:25:42.320 Brylle Girang: I need to create an SOE for… oh, no. I need to create a cursor.
241 00:25:42.670 ⇒ 00:25:43.730 Brylle Girang: amend.
242 00:25:44.450 ⇒ 00:25:46.210 Brylle Girang: Update Notion.
243 00:25:48.530 ⇒ 00:25:49.300 Brylle Girang: So, if I just.
244 00:25:49.300 ⇒ 00:25:50.400 Amber Lin: So, you know.
245 00:25:50.400 ⇒ 00:25:52.500 Brylle Girang: You used agent for this?
246 00:25:52.670 ⇒ 00:26:02.210 Brylle Girang: it will suck us, because it will not… it won’t be able to do anything really helpful, because we didn’t give it enough context to work on, right?
247 00:26:02.210 ⇒ 00:26:07.970 Amber Lin: Oh, so when you use plan, they kind of prompt you for, hey, I need this, I need that.
248 00:26:08.390 ⇒ 00:26:14.199 Brylle Girang: Yeah, so it thinks thoroughly, and it makes sure that it reads everything first.
249 00:26:14.560 ⇒ 00:26:15.170 Amber Lin: Okay.
250 00:26:15.170 ⇒ 00:26:25.180 Brylle Girang: sends you that plan, and then you’ll… you’ll have the option to, like, hey, this is not what I want, this is what I want to do, etc. So it’s like creating a PRD,
251 00:26:25.380 ⇒ 00:26:26.560 Brylle Girang: First.
252 00:26:26.560 ⇒ 00:26:28.009 Amber Lin: Oh, it’s a PRD?
253 00:26:28.010 ⇒ 00:26:35.250 Brylle Girang: A PRD is, I forgot the R, but it’s a product research document, so it’s basically a plan…
254 00:26:35.250 ⇒ 00:26:36.250 Amber Lin: Oh, okay.
255 00:26:36.250 ⇒ 00:26:37.359 Brylle Girang: Oop, thanks.
256 00:26:37.560 ⇒ 00:26:41.429 Brylle Girang: Right, so before building things, you need a PRD, and it’s like.
257 00:26:42.280 ⇒ 00:26:45.559 Brylle Girang: It’s like the planning phase for Cursor, and you can see.
258 00:26:45.560 ⇒ 00:26:46.480 Amber Lin: I see.
259 00:26:46.480 ⇒ 00:26:49.029 Brylle Girang: It even asks questions. So, what is the.
260 00:26:49.030 ⇒ 00:26:49.630 Amber Lin: Yeah.
261 00:26:49.630 ⇒ 00:26:50.909 Brylle Girang: case.
262 00:26:50.910 ⇒ 00:26:58.030 Amber Lin: Wait, can… I’m gonna take a screenshot. I actually want this for our ABC product for the team.
263 00:26:58.160 ⇒ 00:27:02.880 Amber Lin: Because I would like this to ask… like it to ask some questions.
264 00:27:03.610 ⇒ 00:27:05.680 Amber Lin: Great.
265 00:27:06.160 ⇒ 00:27:07.219 Brylle Girang: And then,
266 00:27:07.440 ⇒ 00:27:14.309 Brylle Girang: the output for this will be a plan document. So a plan document is like a step-by-step guide for.
267 00:27:15.420 ⇒ 00:27:27.760 Brylle Girang: So here is where the iteration should come to place, right? It tells us the goal, it tells us the command file, it tells us the behavior, etc. This has been created from, like, one
268 00:27:28.000 ⇒ 00:27:37.649 Brylle Girang: sentence, so I’m not expecting that much, but this is already way above the one-sentence command that we initially asked, right?
269 00:27:38.400 ⇒ 00:27:46.629 Brylle Girang: which is possible, since it asks me questions, like, what’s the use case, right, or sync content from Vault or chat into a Notion page.
270 00:27:46.790 ⇒ 00:27:51.830 Brylle Girang: To also use that answer into, like, refining the actual plan.
271 00:27:51.830 ⇒ 00:27:58.929 Amber Lin: So, is this gonna become a command, like the audit command you showed me, or it… how…
272 00:27:59.280 ⇒ 00:28:01.009 Amber Lin: How is this?
273 00:28:01.630 ⇒ 00:28:11.169 Brylle Girang: It depends on what you ask a cursor to do. Since I asked Cursor to create a cursor command, it will create a command for me, but let’s say, hey.
274 00:28:11.790 ⇒ 00:28:19.120 Brylle Girang: I need you to build an SOP, etc, then it will just build whatever you tell them, and you can confirm that.
275 00:28:19.360 ⇒ 00:28:20.770 Brylle Girang: by…
276 00:28:21.290 ⇒ 00:28:30.009 Brylle Girang: So there’s no output here, but the goal here is to just create a single cursor command that tries to sync content, etc, etc.
277 00:28:30.220 ⇒ 00:28:37.330 Brylle Girang: So, as you can see here, it already used the information available, which is Notion MCP.
278 00:28:37.640 ⇒ 00:28:38.539 Brylle Girang: So it tells us.
279 00:28:38.540 ⇒ 00:28:39.930 Amber Lin: Hmm, I see.
280 00:28:40.210 ⇒ 00:28:46.420 Brylle Girang: Can you tell me a bit more about the MCPs? Like, what MCPs are we connected to, and, like.
281 00:28:46.480 ⇒ 00:28:51.639 Amber Lin: Do I have to connect them, or are they connected? How does that work?
282 00:28:51.640 ⇒ 00:29:03.929 Brylle Girang: Yeah, no worries. So, since the MCPs that we have are connected to our GitHub repo, as long as you have the most… the latest version, rather, then you should be good. You can double.
283 00:29:03.930 ⇒ 00:29:14.829 Amber Lin: How can we connect it to the… to the repo? Because I’m curious. I want to build some tools myself, but I just have no clue how we made it work.
284 00:29:15.050 ⇒ 00:29:17.590 Brylle Girang: Yeah, there are…
285 00:29:17.900 ⇒ 00:29:23.740 Brylle Girang: already a lot of available NCPs in the market, so what we’re just doing is we’re updating this.
286 00:29:24.030 ⇒ 00:29:27.040 Amber Lin: Oh, okay, okay, very cool.
287 00:29:27.040 ⇒ 00:29:35.109 Brylle Girang: Yeah, just adding the MCP servers. I think the recent MCPs that we connected is Notion, and then operating.
288 00:29:36.500 ⇒ 00:29:39.989 Amber Lin: Awesome. Are you aware of what an MCP is?
289 00:29:40.250 ⇒ 00:29:48.459 Amber Lin: It kind of, yeah, I think I reposted someone’s article when it comes out. It’s more like a standardized connector between.
290 00:29:48.460 ⇒ 00:29:50.830 Brylle Girang: Yeah. For the tools to translate.
291 00:29:50.950 ⇒ 00:29:51.720 Amber Lin: Yes.
292 00:29:51.720 ⇒ 00:29:52.450 Brylle Girang: Exactly.
293 00:29:52.470 ⇒ 00:29:58.020 Amber Lin: Yeah. So, are we… sorry, sorry, go ahead. I have questions, but I want to hear what you have to say.
294 00:29:58.160 ⇒ 00:30:08.409 Brylle Girang: No worries, so we have APIs for apps, like, and then APIs are, like, the languages of these apps. It’s how they talk to each other.
295 00:30:09.330 ⇒ 00:30:20.510 Brylle Girang: Then, when AIs were developed, MCPs have also been developed, and these MCPs are, like, the translation services for the APIs over to our agents.
296 00:30:21.290 ⇒ 00:30:29.080 Amber Lin: Okay. So, if we connect to MCP, do we still need to connect APIs, or is this just automatically done? Yeah.
297 00:30:29.080 ⇒ 00:30:30.489 Brylle Girang: It’s already automatic, so…
298 00:30:30.490 ⇒ 00:30:31.050 Amber Lin: I don’t see.
299 00:30:31.050 ⇒ 00:30:36.469 Brylle Girang: does is it uses the API for us, so that we’re not, we’re not the ones.
300 00:30:36.690 ⇒ 00:30:40.500 Amber Lin: I see, I see. So we don’t need to standardize the connection.
301 00:30:40.730 ⇒ 00:30:44.110 Amber Lin: Okay. Questions on…
302 00:30:44.980 ⇒ 00:30:56.740 Amber Lin: I don’t need Figma, but it’s interesting. So what about operating? Are we able to update the hours there, or is that the next command that you were gonna show me?
303 00:30:56.740 ⇒ 00:31:02.010 Brylle Girang: Yeah, exactly. That’s not the next command, but we have…
304 00:31:02.460 ⇒ 00:31:10.419 Brylle Girang: since we have connected the MCP, I also asked the operations team to, like, update the learning materials that we have here, but.
305 00:31:10.950 ⇒ 00:31:20.140 Brylle Girang: Mariah, this is read and write, so it means that it can capture information from operating, and it can also write information over to operating.
306 00:31:20.140 ⇒ 00:31:20.820 Amber Lin: Okay.
307 00:31:20.820 ⇒ 00:31:34.519 Brylle Girang: Great question, yes, it can update the hours for us. And it does so much more, right? For example, in this… when I was testing the MCP, I basically asked Cursor to do all the stuff. For Magic Spoon.
308 00:31:35.320 ⇒ 00:31:37.260 Brylle Girang: Allocate hours.
309 00:31:37.260 ⇒ 00:31:50.200 Amber Lin: I see. Can you actually ask it to give us a summary of what’s currently in… in operating? I want to see, like, if it’s accurate about active projects, how it presents the allocations…
310 00:31:50.400 ⇒ 00:31:53.820 Amber Lin: And then we can also see if there’s something we need to update.
311 00:31:54.030 ⇒ 00:31:57.019 Brylle Girang: Summary of Project Square.
312 00:31:57.580 ⇒ 00:32:00.759 Brylle Girang: Amber is the owner, or VP.
313 00:32:02.650 ⇒ 00:32:08.010 Amber Lin: I don’t think I owe money projects. I just wanted to see, like, projects overall.
314 00:32:08.680 ⇒ 00:32:18.719 Brylle Girang: It should be able to do that, and this is also one thing that I asked Rico to utilize, so that when he creates the operating reports, he doesn’t need to go through the site.
315 00:32:18.870 ⇒ 00:32:19.440 Brylle Girang: And then…
316 00:32:19.440 ⇒ 00:32:20.259 Amber Lin: I see.
317 00:32:20.260 ⇒ 00:32:21.060 Brylle Girang: gratefully.
318 00:32:26.110 ⇒ 00:32:29.839 Brylle Girang: So it might be a little slow, because we haven’t created
319 00:32:29.980 ⇒ 00:32:37.740 Brylle Girang: Operating specific skills yet. But once we have the skills ready, then it should be a bit faster.
320 00:32:43.010 ⇒ 00:32:47.189 Brylle Girang: Okay, so it just… it’s just asking me to authenticate. Let me do that one.
321 00:33:14.750 ⇒ 00:33:23.900 Brylle Girang: While we’re waiting for this, you can double-check the available MCPs by going to Cursor, Settings, and then Cursor Settings.
322 00:33:25.260 ⇒ 00:33:27.789 Brylle Girang: Let’s go to Tools and MCP.
323 00:33:29.230 ⇒ 00:33:31.469 Amber Lin: Oh… okay.
324 00:33:31.930 ⇒ 00:33:41.560 Amber Lin: Very cool. So once I connect to the… oh, it’s not connected. I see. So once I connect to our Brainforge platform, I should be able to see that.
325 00:33:41.810 ⇒ 00:33:42.520 Brylle Girang: Yes.
326 00:33:42.700 ⇒ 00:33:43.890 Amber Lin: Awesome, okay.
327 00:33:43.890 ⇒ 00:33:46.899 Brylle Girang: authentication failed. Let me try again.
328 00:33:50.800 ⇒ 00:33:57.960 Brylle Girang: Okay, so it’s taking this long because the MCP was not connected properly. Let me just refresh this one.
329 00:34:09.150 ⇒ 00:34:14.320 Brylle Girang: Okay, while we’re waiting for this, then let’s go back. I just wanted to… .
330 00:34:14.320 ⇒ 00:34:17.929 Amber Lin: Show to you the second command that we have created, so it’s just…
331 00:34:17.929 ⇒ 00:34:19.540 Brylle Girang: The weekly kickoff.
332 00:34:20.679 ⇒ 00:34:37.399 Brylle Girang: So this is what we have been using every Mondays to, like, get us up to speed for the clients and, like, provide them a quick, weekly kickoff summary. If you remember, I sent over this message to the Element channel, where I asked for your feedback.
333 00:34:37.609 ⇒ 00:34:38.480 Amber Lin: I think.
334 00:34:38.489 ⇒ 00:34:39.949 Brylle Girang: via this command.
335 00:34:42.310 ⇒ 00:34:50.050 Amber Lin: Wow. I mean, I saw your update, it was really good. It took me so much time when I was doing project management to…
336 00:34:50.179 ⇒ 00:34:57.180 Amber Lin: Like, when I was not doing work on a project to summarize everything, because I didn’t know what was going on, so this is really cool.
337 00:34:57.510 ⇒ 00:35:07.049 Brylle Girang: Yeah, exactly, and that’s the beauty of our vault right now, because everything’s there. You can just ask Cursor, and then it will be able to help you get up to speed.
338 00:35:07.200 ⇒ 00:35:13.080 Brylle Girang: Even if you were not in the meetings, even if you were not, you know, if you were… if you were absent for 3 days.
339 00:35:13.180 ⇒ 00:35:32.239 Brylle Girang: then you’ll be able to get up to speed. So what it does is it just gets the transcripts from the last 7 days, since this is based on the Monday start date. So last 7 days, what happened, what is in our linear project. In the future, hopefully, how does our timeline look like? Are we on track?
340 00:35:32.340 ⇒ 00:35:38.290 Brylle Girang: what’s the… what’s the project called? Echatra? And then it will create… A summary for you.
341 00:35:39.210 ⇒ 00:35:40.160 Amber Lin: I see.
342 00:35:42.870 ⇒ 00:35:43.740 Brylle Girang: So, there.
343 00:35:46.580 ⇒ 00:36:01.229 Brylle Girang: This one. So, this is in a markdown file, because we wanted to, like, get the logs of the weekly updates, so that it can also be referenced by cursor the next week, and cross-check if there’s anything missing, etc.
344 00:36:02.050 ⇒ 00:36:04.090 Brylle Girang: Something that you can do every week.
345 00:36:04.400 ⇒ 00:36:08.209 Amber Lin: And something that we have been pushing our CSOs to do.
346 00:36:08.780 ⇒ 00:36:09.540 Amber Lin: Okay.
347 00:36:09.540 ⇒ 00:36:12.069 Brylle Girang: Back to the project summary, so…
348 00:36:15.990 ⇒ 00:36:17.639 Brylle Girang: Oh, it still failed.
349 00:36:18.120 ⇒ 00:36:21.890 Amber Lin: Yeah, I think it pulled from other stuff.
350 00:36:22.530 ⇒ 00:36:26.590 Brylle Girang: Yeah, it tells us that the operating MCP requires authentication, so…
351 00:36:26.620 ⇒ 00:36:34.829 Amber Lin: I see, I see, okay. I mean, also, MinuteMD is AMBLE, so I think if we connect, it might be even more accurate.
352 00:36:44.150 ⇒ 00:36:45.319 Brylle Girang: Let me try.
353 00:36:50.510 ⇒ 00:36:55.779 Brylle Girang: So, the MCP for operating is in the alpha state right now, they just activated this last one.
354 00:36:56.360 ⇒ 00:36:58.089 Brylle Girang: So, this is expensive.
355 00:36:58.090 ⇒ 00:37:05.850 Amber Lin: Gotcha. Okay, okay. That’s okay. I mean, I don’t need that immediately, but that will be awesome.
356 00:37:06.860 ⇒ 00:37:14.180 Brylle Girang: When, when we connotes us to, like, do stuff with operating, try this out. It’ll be a lot easier.
357 00:37:14.580 ⇒ 00:37:22.639 Brylle Girang: I was… I was having a hard time trying to, like, calculate and balance the allocations per team member, so I just asked her, sir, hey.
358 00:37:23.110 ⇒ 00:37:25.639 Brylle Girang: That’s how the allocation, etc, and did it.
359 00:37:26.230 ⇒ 00:37:29.840 Brylle Girang: It automatically calculated the percentage.
360 00:37:33.660 ⇒ 00:37:36.319 Brylle Girang: It’s failing right now, I’m not sure why.
361 00:37:40.020 ⇒ 00:37:53.559 Brylle Girang: So let me give this feedback to the operating team. But yeah, that’s the… that’s, like, the idea, where we can… we can just use cursor to update operating, make sure that we are always on track when it comes to the allocations.
362 00:37:54.830 ⇒ 00:37:57.909 Brylle Girang: Yeah, I think that’s everything that I can share.
363 00:38:00.570 ⇒ 00:38:17.560 Amber Lin: Yeah, this is really, really helpful. I mean, I’ve not used it for EP tests at all, so I think on my side, my action items is to, set up the GitHub first. Has Rico got back to you?
364 00:38:19.300 ⇒ 00:38:26.259 Brylle Girang: Not yet, but I’m going to follow up, but yeah, set up your cursor account, make sure that you’re in our enterprise account.
365 00:38:26.430 ⇒ 00:38:27.200 Brylle Girang: And then…
366 00:38:27.200 ⇒ 00:38:32.319 Amber Lin: Yeah, I’m gonna go… I’ll send an update in the AI team. Okay.
367 00:38:35.180 ⇒ 00:38:39.530 Brylle Girang: I’m also working on, like, training materials for how to set up.
368 00:38:40.410 ⇒ 00:38:42.849 Brylle Girang: cursor in GitHub, so…
369 00:38:43.620 ⇒ 00:38:47.340 Brylle Girang: I’m going to make sure that I share that with the whole team, just in case you need it.
370 00:38:49.550 ⇒ 00:38:50.480 Amber Lin: Gotcha.
371 00:38:53.780 ⇒ 00:39:13.370 Brylle Girang: Yeah, you know, please, please utilize Cursor, and if you think of ways on how you can use Cursor to, like, speed up your daily workflow, that would be amazing, especially when it comes to EP work, right? This EP work are mostly redundant, and those are the stuff that we want to
372 00:39:13.530 ⇒ 00:39:19.020 Brylle Girang: automate using AI so that you can focus on the amazing analytics that you’re doing.
373 00:39:19.470 ⇒ 00:39:26.170 Brylle Girang: And once we have enough info, then maybe we can dive into what you’re doing when it comes to analytics in the future.
374 00:39:26.580 ⇒ 00:39:29.880 Brylle Girang: See how we can improve your workflows. I think…
375 00:39:29.880 ⇒ 00:39:42.379 Amber Lin: And I know this is something we want to do. I also want to help us write down how I do it, because especially for the easier analysis, it’s really not that hard. So…
376 00:39:42.850 ⇒ 00:39:48.129 Amber Lin: I think that’s something that we can start writing out and creating process steps.
377 00:39:48.570 ⇒ 00:39:49.690 Brylle Girang: Yeah, yeah.
378 00:39:49.970 ⇒ 00:39:53.569 Brylle Girang: That skills would be really helpful, because,
379 00:39:53.940 ⇒ 00:40:05.800 Brylle Girang: how skills work in Cursor is that it just gives the AI a bit more info up front, so that it doesn’t go all over the place, just to try to understand what to do. So…
380 00:40:05.800 ⇒ 00:40:10.070 Amber Lin: When you say skills, what does that mean?
381 00:40:10.510 ⇒ 00:40:17.280 Amber Lin: So, in Cursor and in most AI agents, AI platforms, they have, like, 3.
382 00:40:17.280 ⇒ 00:40:27.780 Brylle Girang: predefined parameters. So we have the commands, which I just showed you. So the commands are like, if I tell you this, do this.
383 00:40:28.220 ⇒ 00:40:30.430 Amber Lin: So it’s a workflow.
384 00:40:30.430 ⇒ 00:40:46.249 Brylle Girang: Yeah, it’s a workflow. Exactly. The second one are rules. So, rules are just do’s and don’ts. Don’t do this. If the agent… if the user tells you to look at operating, don’t look at Notion, etc.
385 00:40:46.870 ⇒ 00:40:52.810 Brylle Girang: So, those are… well, as the name suggests, those are actually rules, actual rules for the agent.
386 00:40:53.100 ⇒ 00:40:55.639 Brylle Girang: Now, the third one are skills.
387 00:40:55.900 ⇒ 00:40:59.550 Brylle Girang: So, skills are, like, SOPs.
388 00:40:59.690 ⇒ 00:41:07.560 Brylle Girang: For the agents, but its main purpose is to ensure that the agents’ focus do not go out of bounds.
389 00:41:07.760 ⇒ 00:41:13.269 Brylle Girang: So, when we tell… when we… when we tell the agent to, like, hey.
390 00:41:13.360 ⇒ 00:41:29.800 Brylle Girang: look at my operating projects. You see earlier, it looked at everything. It looked at linear, it looked at the platform, it looked at the vault, and that’s because we don’t have the skills yet, or a cursor. If we have a skill telling them that, hey,
391 00:41:30.250 ⇒ 00:41:39.729 Brylle Girang: To look at operating projects, use the operating MCP, use these specific JSON files, use this specific trigger, etc.
392 00:41:39.930 ⇒ 00:41:47.599 Brylle Girang: Then, if the agent sees that instruction, it will forget everything and just focus on that one path.
393 00:41:48.380 ⇒ 00:42:01.109 Brylle Girang: Okay. So, why is that helpful? It keeps costs down, because tokens are not burnt as much as what could have been if there are no skills. And two.
394 00:42:01.490 ⇒ 00:42:05.069 Brylle Girang: It makes the prompt faster, because it doesn’t, like.
395 00:42:05.070 ⇒ 00:42:11.350 Amber Lin: How does it differ from a command? Because they’re all, like, SOP-ish.
396 00:42:11.750 ⇒ 00:42:12.960 Amber Lin: Thanks.
397 00:42:12.960 ⇒ 00:42:20.460 Brylle Girang: Yeah, I was… I was also confused at first, but commands are linked… are actually linked to skills.
398 00:42:20.670 ⇒ 00:42:25.039 Brylle Girang: Like, commands are the step-by-steps. It can use
399 00:42:25.230 ⇒ 00:42:28.379 Brylle Girang: So many skills in the same command.
400 00:42:28.540 ⇒ 00:42:36.630 Brylle Girang: So, the command can be, hey, do the linear audit. It uses skill 1, look at linear projects, it uses skill 2.
401 00:42:37.100 ⇒ 00:42:37.599 Amber Lin: Like the crap.
402 00:42:37.600 ⇒ 00:42:40.120 Brylle Girang: transcripts. It uses QL3.
403 00:42:40.440 ⇒ 00:42:49.729 Brylle Girang: check the vault. And if you have a rule telling that, hey, only look at the Brainforge platform, then it will not look at any other vaults outside of that.
404 00:42:50.120 ⇒ 00:42:51.640 Brylle Girang: Does that make sense?
405 00:42:51.640 ⇒ 00:43:07.329 Amber Lin: Yeah, I see, I see, that makes sense. Very cool. Okay, I don’t have any more questions for now. I think once, once Rico gives me access, I’ll connect it, and then I’ll… I’ll attempt those two, commands.
406 00:43:07.330 ⇒ 00:43:08.070 Brylle Girang: Okay.
407 00:43:08.210 ⇒ 00:43:10.550 Brylle Girang: Thank you so much, Amber, and then maybe…
408 00:43:10.550 ⇒ 00:43:14.769 Amber Lin: Thank you so much for walking me through, you’re very patient, and this was really, really helpful.
409 00:43:14.770 ⇒ 00:43:23.419 Brylle Girang: Of course, and if we, you know, if we have the chance, maybe let’s regroup after 2 weeks, see where we are at, and then let’s see how we can help.
410 00:43:23.420 ⇒ 00:43:28.439 Amber Lin: Yeah, yeah, feel free to, like, put another, like, 30 minutes, and we can talk again.
411 00:43:28.440 ⇒ 00:43:30.539 Brylle Girang: Okay. Thank you so much, bye-bye.
412 00:43:30.540 ⇒ 00:43:32.470 Amber Lin: Yeah, thank you, bye!