WEBVTT

1
00:00:09.530 --> 00:00:10.600
Tom Babb: We are live.

2
00:00:25.610 --> 00:00:26.300
Derek Peters: Right.

3
00:00:27.050 --> 00:00:28.000
Derek Peters: Right.

4
00:00:28.700 --> 00:00:35.290
Derek Peters: We'll give, we'll give folks a few moments to join. I appreciate you being here. We'll get started in just a second.

5
00:00:35.710 --> 00:00:44.390
Derek Peters: I'm gonna post some stuff in the chat, just to get some interaction while we get, as some people are getting in, we're gonna start shortly. I look forward to,

6
00:00:44.870 --> 00:00:48.210
Derek Peters: to going over this lecture with you. Welcome to night school.

7
00:01:04.040 --> 00:01:12.260
Derek Peters: All right, in the chat, you can also put up, put where you're calling from. It's always neat to see where everybody's from and all the different places that it's reaching out to.

8
00:01:12.600 --> 00:01:17.640
Derek Peters: All right, Carl's from Austin, we got some San Francisco area, Nashville.

9
00:01:18.090 --> 00:01:19.320
Derek Peters: Go Rocky Talk!

10
00:01:20.410 --> 00:01:21.600
Derek Peters: Orlando?

11
00:01:22.350 --> 00:01:26.139
Derek Peters: Lincoln, not… oh, gosh, Philippines.

12
00:01:29.970 --> 00:01:32.460
Derek Peters: Dubai. All around the world!

13
00:01:35.370 --> 00:01:41.370
Derek Peters: Got Maryland, got Virginia, nice. I'm from Tennessee, so I think I saw somebody in from Nashville.

14
00:01:43.880 --> 00:01:46.680
Derek Peters: I do have a little draw, so I'm gonna try…

15
00:01:47.280 --> 00:01:49.070
Derek Peters: To not let it get the best of me.

16
00:01:51.770 --> 00:01:52.960
Derek Peters: Alright.

17
00:01:57.750 --> 00:02:03.519
Derek Peters: Brooklyn Alright, while you guys are answering that question…

18
00:02:03.710 --> 00:02:13.720
Derek Peters: We'll go ahead and get started. So, it looks like we've got a good crowd with us. I want to start by introducing myself. My name is Derek Peters.

19
00:02:13.900 --> 00:02:17.950
Derek Peters: I work with the product engineering team here at Gauntlet.

20
00:02:18.090 --> 00:02:23.810
Derek Peters: I have 15 years in engineering, planning, and go-to-market, experience.

21
00:02:24.190 --> 00:02:31.849
Derek Peters: and I'm from Tennessee, and one of my biggest hobbies is I love to go fly fishing.

22
00:02:32.390 --> 00:02:38.270
Derek Peters: So, today, and what brings you here today is, for tonight's night school.

23
00:02:39.050 --> 00:02:44.529
Derek Peters: I want to talk about… kind of take a step back from coding, developing, and…

24
00:02:45.020 --> 00:02:49.930
Derek Peters: Everything that we're willing as engineers to just jump right into the mix.

25
00:02:50.580 --> 00:02:56.070
Derek Peters: And just take a step back and realize that you're not just building it, you're getting it launched.

26
00:02:56.660 --> 00:02:59.830
Derek Peters: And… What does that mean?

27
00:03:00.800 --> 00:03:04.279
Derek Peters: So that's exactly what we're going to talk about tonight.

28
00:03:04.730 --> 00:03:10.700
Derek Peters: So… Let's go ahead and go to the next slide.

29
00:03:11.070 --> 00:03:14.149
Derek Peters: So, here's the shape of the next hour.

30
00:03:15.480 --> 00:03:22.489
Derek Peters: We open with why building products just changed, and the methodology that we followed.

31
00:03:23.140 --> 00:03:26.539
Derek Peters: Then, we'll take a look at 3 important pieces.

32
00:03:26.820 --> 00:03:29.290
Derek Peters: All in one real product.

33
00:03:29.910 --> 00:03:32.490
Derek Peters: So, we're gonna touch base on planning.

34
00:03:32.760 --> 00:03:42.870
Derek Peters: So that's turning an idea into a vision document, a product requirements document, also referred to as a PRD, and then a roadmap

35
00:03:43.160 --> 00:03:45.659
Derek Peters: the AI could actually build from.

36
00:03:47.380 --> 00:03:49.289
Derek Peters: And then we'll talk about prompting.

37
00:03:51.290 --> 00:04:01.149
Derek Peters: you are only as good as your prompt, so I feel like now, like, it's a mix between prompt engineering and engineering. Prompting is very important.

38
00:04:01.430 --> 00:04:04.160
Derek Peters: As you steer where your product is going.

39
00:04:04.970 --> 00:04:12.460
Derek Peters: So you're driving the build straight out of those documents. And then we'll talk about… we'll touch base on shipping.

40
00:04:12.850 --> 00:04:16.769
Derek Peters: Release. How do we measure it and close the loop?

41
00:04:17.180 --> 00:04:26.170
Derek Peters: And then, we'll also have some time for Q&A. We're also monitoring… oh, I just saw somebody from Denver, nice.

42
00:04:26.570 --> 00:04:31.479
Derek Peters: We're… ADD moment. But by the end, you'll see exactly which

43
00:04:31.620 --> 00:04:36.150
Derek Peters: Of these three are often skipped, and why they need to go together.

44
00:04:38.320 --> 00:04:44.219
Derek Peters: And I'd bet money it was one that you may have bookend and just forgot about.

45
00:04:44.590 --> 00:04:46.040
Derek Peters: So let's dive in.

46
00:04:47.170 --> 00:04:54.359
Derek Peters: So we're gonna start here, because it reframes everything else. So, for decades.

47
00:04:55.030 --> 00:04:59.809
Derek Peters: The bottleneck in building software was typing.

48
00:05:00.880 --> 00:05:07.770
Derek Peters: was coding, was executing. Turning a decision into working code, took weeks.

49
00:05:08.470 --> 00:05:13.069
Derek Peters: Well, now, with all the great technologies that are out there, That's gone.

50
00:05:14.090 --> 00:05:16.039
Derek Peters: So look at these two bars.

51
00:05:16.280 --> 00:05:23.670
Derek Peters: It's the same timeframe. It's the same week, or sprint, or however you operate. It's the same team.

52
00:05:24.250 --> 00:05:25.849
Derek Peters: So, the old way.

53
00:05:26.180 --> 00:05:32.560
Derek Peters: That whole week brought you one shift feature, or that whole sprint brought you one shipped feature.

54
00:05:32.880 --> 00:05:36.210
Derek Peters: The new way, the typing collapses.

55
00:05:36.770 --> 00:05:42.170
Derek Peters: So, in the very same week you ship, you can ship, let's say, 3 features.

56
00:05:42.360 --> 00:05:50.000
Derek Peters: three times the amount of work. In some cases, four and five. It depends on how good you are at planning.

57
00:05:50.650 --> 00:05:59.930
Derek Peters: And hopefully this will help you become a better planner and separate vibe coding from being a product

58
00:06:00.690 --> 00:06:06.709
Derek Peters: Have some product sense, and kind of go into… That development phase.

59
00:06:08.370 --> 00:06:20.190
Derek Peters: So, it's being ruthless and planning before, and then shipping after. So that's the line between what I mentioned before, VOB coding, let's say VOB coding a slick demo.

60
00:06:20.300 --> 00:06:25.980
Derek Peters: And actually building a product That users are going to utilize.

61
00:06:32.600 --> 00:06:35.790
Derek Peters: So here's the methodology that we're going to use tonight.

62
00:06:36.370 --> 00:06:38.989
Derek Peters: I call it the AI-first product loop.

63
00:06:39.450 --> 00:06:50.359
Derek Peters: And it's just 5 moves. You frame the vision, that's your product vision document, You define the requirements.

64
00:06:54.150 --> 00:06:59.669
Derek Peters: And that's your PRD. You sequence the work, that's your roadmap build.

65
00:06:59.860 --> 00:07:08.049
Derek Peters: And then you go into building. That's where you finally prompt the AI and work on shipping those incremental releases.

66
00:07:08.510 --> 00:07:18.110
Derek Peters: And at that point, you're in… you release and you measure. But the most important part of this whole diagram is the arrow that loops back.

67
00:07:19.260 --> 00:07:27.010
Derek Peters: So… So everything that you learn in production feeds to the next vision.

68
00:07:27.120 --> 00:07:29.870
Derek Peters: So notice the mapping along the bottom frame.

69
00:07:30.330 --> 00:07:32.150
Derek Peters: Define the sequence.

70
00:07:33.220 --> 00:07:39.479
Derek Peters: well, yeah, define and sync what's all that lives inside of planning. The first 3 to 5 moves

71
00:07:40.270 --> 00:07:42.829
Derek Peters: Spent before you write a single prompt?

72
00:07:42.970 --> 00:07:47.799
Derek Peters: And… So, prompting is one move, shipping is one move.

73
00:07:48.000 --> 00:07:56.990
Derek Peters: plus the loop home, so it kind of goes in full circle. Keep that picture in your head, because every slide that we go through, everything that we discuss tonight.

74
00:07:57.270 --> 00:08:04.570
Derek Peters: is going to be… it's going to come up, and it's going to frame together nicely. And it's a loop, it's not a line.

75
00:08:05.060 --> 00:08:11.020
Derek Peters: So, Learn feeds straight back into frame, so your second version is always sharper than your first.

76
00:08:11.510 --> 00:08:17.909
Derek Peters: And then pretend you could get, you know, the whole thing at once. This assumes you won't.

77
00:08:18.080 --> 00:08:24.940
Derek Peters: And the builds are… as we get into the building piece, we'll kind of get into that cadence, so…

78
00:08:25.570 --> 00:08:27.440
Derek Peters: Let's talk about…

79
00:08:28.430 --> 00:08:42.679
Derek Peters: making it real. So, I'm gonna use, you know, a product that I'm gonna reference back to. It's a fictional product, but it is a real product that could be created. So, tonight's product is Cadence.

80
00:08:43.049 --> 00:08:46.720
Derek Peters: It's a task tracker with AI built in.

81
00:08:47.220 --> 00:08:51.179
Derek Peters: So, think of it as, like, you pasted all this chaos in your head.

82
00:08:51.490 --> 00:08:55.319
Derek Peters: The half-thoughts, the… everything that you want to accomplish.

83
00:08:55.740 --> 00:09:03.700
Derek Peters: All the to-dos that you have to do across notes and chats and everything that you have to do in, you know, your day-to-day work environment.

84
00:09:05.880 --> 00:09:09.550
Derek Peters: And cadence turns it into a clean, prioritized list in seconds.

85
00:09:09.880 --> 00:09:19.150
Derek Peters: Now, I picked it for a couple reasons. Everyone lives this pain every day. Your tasks are spread across notes, messages, maybe some post-it notes.

86
00:09:19.560 --> 00:09:22.379
Derek Peters: And your own memory, and let's say.

87
00:09:22.950 --> 00:09:28.640
Derek Peters: Yeah, they exist here, and sometimes you forget them because you get sidetracked, or you see some squirrels.

88
00:09:28.900 --> 00:09:29.840
Derek Peters: But…

89
00:09:33.280 --> 00:09:35.130
Derek Peters: It's hard to keep up with.

90
00:09:35.290 --> 00:09:39.129
Derek Peters: So I picked that because it's a real pain point for a lot of us.

91
00:09:39.300 --> 00:09:52.639
Derek Peters: And two, there's some real depth into this. You're asking a model to read a messy human input, and then hand back something structured and trustworthy, which is exactly where AI products get hard.

92
00:09:53.520 --> 00:10:01.100
Derek Peters: And there's an end state on the right. That's a brain dump in, prioritize task list out.

93
00:10:01.510 --> 00:10:02.580
Derek Peters: And…

94
00:10:02.790 --> 00:10:09.919
Derek Peters: as I'm planning this, or taking, like… and I'm gonna take you through the steps that I would go through in planning this.

95
00:10:11.230 --> 00:10:20.520
Derek Peters: But we'll trace the first feature from ideation to ship. And because every prompt and every result is right here on the slides.

96
00:10:21.980 --> 00:10:29.759
Derek Peters: the experience that I want you to get is understanding the workflow, and… and the reason why I didn't want to make it a live exercise is

97
00:10:29.950 --> 00:10:33.439
Derek Peters: It's difficult to follow watching others type, so…

98
00:10:33.840 --> 00:10:39.800
Derek Peters: Let's talk, you know, before we touch, like, a single tool, or talk about a single tool.

99
00:10:39.960 --> 00:10:43.980
Derek Peters: The most important choice of the night.

100
00:10:45.040 --> 00:10:47.309
Derek Peters: There's two ways to build with AI.

101
00:10:48.090 --> 00:10:50.309
Derek Peters: And they start with the same prompt.

102
00:10:53.380 --> 00:10:56.260
Derek Peters: But they end completely different places.

103
00:10:56.930 --> 00:11:00.080
Derek Peters: So, on the left is VOB coding.

104
00:11:00.260 --> 00:11:03.460
Derek Peters: You open chat, you type, build me a task app.

105
00:11:03.610 --> 00:11:05.580
Derek Peters: And chase whatever comes back.

106
00:11:06.320 --> 00:11:11.079
Derek Peters: Your prompt after prompt, You're burning tokens.

107
00:11:12.220 --> 00:11:17.270
Derek Peters: It's intoxicating, it's draining, and while it is fast.

108
00:11:17.740 --> 00:11:30.449
Derek Peters: really all you're getting is a… is a demo. There wasn't a whole lot of thought, a whole lot of planning behind it. And the problem is, you end up somewhere that you can't explain. And this is where I find engineers

109
00:11:31.000 --> 00:11:41.690
Derek Peters: struggle sometimes, because it is so easy to… like, with all the tools that are out there, it's so easy to just get in there to start typing, and not understand the why.

110
00:11:42.440 --> 00:11:45.329
Derek Peters: So, you gotta be the master of the ship as you're coding.

111
00:11:46.580 --> 00:11:52.660
Derek Peters: As you're prompting, as you're planning. Planning, to me, is the most important aspect of engineering.

112
00:11:53.940 --> 00:12:00.649
Derek Peters: And you still have to plan. You can use all the resources and tools out there to kind of help bounce ideas off

113
00:12:00.760 --> 00:12:03.249
Derek Peters: And we'll get into that tonight.

114
00:12:03.490 --> 00:12:10.380
Derek Peters: On the right, though, we… you know, is what we actually teach here at Gauntlet.

115
00:12:10.870 --> 00:12:12.989
Derek Peters: Plan, and then prompt.

116
00:12:13.420 --> 00:12:22.100
Derek Peters: you decide… The, you know, you decide what's… what's the best first, you know, what's best going forward.

117
00:12:22.470 --> 00:12:29.790
Derek Peters: And then… You flesh out the vision, you flesh out the requirements, you create a roadmap.

118
00:12:30.340 --> 00:12:34.139
Derek Peters: You create your architecture diagram, and then…

119
00:12:34.420 --> 00:12:36.460
Derek Peters: Once you have all those rich.

120
00:12:37.530 --> 00:12:41.800
Derek Peters: documents going in, then AI can help you

121
00:12:42.270 --> 00:12:46.580
Derek Peters: Build from those documents, and now you can steer.

122
00:12:47.070 --> 00:12:50.740
Derek Peters: And be a product manager, and let…

123
00:12:51.720 --> 00:12:55.190
Derek Peters: The agents be your development team.

124
00:12:55.720 --> 00:13:00.519
Derek Peters: Now, here is probably one of the most common things that people miss.

125
00:13:01.270 --> 00:13:03.480
Derek Peters: It's… it's not slower.

126
00:13:04.350 --> 00:13:14.089
Derek Peters: Same model, same speed, But between Vive Coding and planning, Let's see, is prompting…

127
00:13:19.730 --> 00:13:20.530
Derek Peters: Yeah.

128
00:13:20.850 --> 00:13:27.769
Derek Peters: You can even… you can even bounce some ideas off it and have it create better prompts.

129
00:13:28.680 --> 00:13:29.830
Derek Peters: Absolutely.

130
00:13:30.450 --> 00:13:35.600
Derek Peters: It's the workflow wrapped around it, so… The biggest miss…

131
00:13:37.800 --> 00:13:44.689
Derek Peters: that I usually, or typically see, it, it, well…

132
00:13:45.730 --> 00:13:48.030
Derek Peters: Let's… let's get into talking about it.

133
00:13:51.590 --> 00:13:55.389
Derek Peters: And I also want to talk about what we actually use per stage.

134
00:13:55.620 --> 00:13:59.670
Derek Peters: And I want to talk about tooling, for example, because it is important.

135
00:14:00.120 --> 00:14:04.920
Derek Peters: And it's usually the first thing that people want to, like, dive into and ask.

136
00:14:07.780 --> 00:14:10.239
Derek Peters: And it's the workflow that matters.

137
00:14:10.410 --> 00:14:14.699
Derek Peters: Far more than, let's say, the brand of tooling that we use.

138
00:14:14.960 --> 00:14:20.939
Derek Peters: So, for planning, you won't… A good frontier reasoning model.

139
00:14:21.340 --> 00:14:28.190
Derek Peters: So, like, Claude, Opus, GPT, you can even use Sonnet for just bouncing ideas.

140
00:14:28.460 --> 00:14:35.360
Derek Peters: Gemini, you know, whatever you got handy, those are all great tools to kind of help start for planning.

141
00:14:36.210 --> 00:14:53.389
Derek Peters: And I run it in Cloud Desktop, or sometimes I'll bounce it off just to… I'm always in the mindset, trust but verify I'm not bounce it off a GPT with a product so that my vision and my requirements ride along in every message.

142
00:14:53.600 --> 00:14:59.709
Derek Peters: And then for prompting, this is where you move into an agenic coding tool.

143
00:14:59.990 --> 00:15:10.780
Derek Peters: And there's really two camps right now. So you have Anthropic's Cloud Code, you have Codex, and you do have Cursor.

144
00:15:11.050 --> 00:15:18.630
Derek Peters: But… it's a whole new system. It's a terminal command line interface, so it's a CLI,

145
00:15:18.740 --> 00:15:24.100
Derek Peters: And an extension that lives inside Cursor and VS Code.

146
00:15:24.810 --> 00:15:28.699
Derek Peters: Cursor itself is the editor, a lot of people build in it.

147
00:15:28.980 --> 00:15:36.279
Derek Peters: But, you know, use a strong model. So Claude Sonnet or Opus is good, or GPT 5.5,

148
00:15:36.380 --> 00:15:49.260
Derek Peters: for real building. So, like, whenever you're actually… you get into your building, those three LLMs are really strong for, helping you code and turn those ideas into something that real users can use.

149
00:15:51.790 --> 00:15:59.339
Derek Peters: So, let's go ahead and kick this off with planning. So, this is… this is the first and most important step, in my mind.

150
00:16:04.970 --> 00:16:14.290
Derek Peters: So… I'm going to… Ask you to do something that feels… feels backwards, And let's talk about speed.

151
00:16:15.010 --> 00:16:18.409
Derek Peters: Because in the world of AI-first development.

152
00:16:18.520 --> 00:16:23.490
Derek Peters: It's go faster, go faster, ship, submit, go faster, go faster.

153
00:16:24.790 --> 00:16:28.910
Derek Peters: You're still going fast, but take a step back and slow down.

154
00:16:30.150 --> 00:16:35.149
Derek Peters: The single biggest predictor of whatever an AI build

155
00:16:35.340 --> 00:16:40.470
Derek Peters: Succeeds is how much thinking happened before the chat window ever opened.

156
00:16:41.410 --> 00:16:53.060
Derek Peters: When a build goes sideways, 9 times out of 10, it's because someone asked the model to invent a plan, or to do the planning for them based off of a couple words.

157
00:16:55.250 --> 00:16:56.670
Derek Peters: I'm looking at the chat.

158
00:16:59.360 --> 00:17:06.180
Derek Peters: So, Mark asked, how do you choose between the three agents? One front-end or back-end? Well, so…

159
00:17:06.869 --> 00:17:09.409
Derek Peters: Usually, I'll start with, like, Claude Code.

160
00:17:09.530 --> 00:17:12.839
Derek Peters: And, like, when I'm in my development phase.

161
00:17:13.030 --> 00:17:18.379
Derek Peters: Of what I want to accomplish, and it does a pretty good job with both back-end and front-end.

162
00:17:20.130 --> 00:17:23.500
Derek Peters: But we'll definitely get into that in the shipping piece.

163
00:17:24.609 --> 00:17:27.620
Derek Peters: So, so, for the next few minutes.

164
00:17:29.100 --> 00:17:31.199
Derek Peters: We'll take a look at,

165
00:17:31.570 --> 00:17:34.729
Derek Peters: You know, 3 short documents that…

166
00:17:34.840 --> 00:17:41.730
Derek Peters: will help you be better planners along the way. That's the vision document, requirements document, and a roadmap.

167
00:17:42.690 --> 00:17:43.840
Derek Peters: And…

168
00:17:44.170 --> 00:17:53.560
Derek Peters: Only then do you want to prompt. No code yet. This is just thinking out loud on paper and know what the objection is and what you're trying to accomplish.

169
00:17:55.750 --> 00:17:58.579
Derek Peters: So, you start with a vision document.

170
00:17:59.280 --> 00:18:07.019
Derek Peters: So, before features, before Roadmap, You write out your vision document.

171
00:18:08.240 --> 00:18:12.099
Derek Peters: One page on why this exists, and who's it for?

172
00:18:12.940 --> 00:18:15.289
Derek Peters: What problem are we trying to solve?

173
00:18:17.090 --> 00:18:19.090
Derek Peters: Who's gonna benefit from it?

174
00:18:19.730 --> 00:18:22.340
Derek Peters: And here's the cadence on the right.

175
00:18:22.620 --> 00:18:23.830
Derek Peters: So, the vision.

176
00:18:24.050 --> 00:18:28.490
Derek Peters: Turn the chaos in your head into a clear, prioritized plan.

177
00:18:28.980 --> 00:18:34.820
Derek Peters: So the user, the core problem. Tasks live everywhere, and…

178
00:18:35.130 --> 00:18:39.800
Derek Peters: Prioritizing can sometimes be a constant manual work.

179
00:18:40.300 --> 00:18:42.330
Derek Peters: And the one metric that matters?

180
00:18:42.670 --> 00:18:49.160
Derek Peters: Plan task… planned tasks are actually completed. That's all that matters, is what you plan gets completed.

181
00:18:50.860 --> 00:18:58.849
Derek Peters: And then… and principles, so which, you know, let's say, which, for a tool that you live in all day.

182
00:18:59.560 --> 00:19:06.179
Derek Peters: means calm, trustworthy priority, and you always stay in control. Now, how do you write this with AI?

183
00:19:07.800 --> 00:19:10.439
Derek Peters: You do not ask to invent your vision.

184
00:19:11.090 --> 00:19:16.049
Derek Peters: One thing that I like to do, and I have an example here, is…

185
00:19:18.140 --> 00:19:20.310
Derek Peters: I ask it to interview me.

186
00:19:20.510 --> 00:19:24.579
Derek Peters: if I'm getting blocked in my mind, like… and look…

187
00:19:25.390 --> 00:19:28.819
Derek Peters: Kind of look how the prompt is built, like, act as a product coach.

188
00:19:29.000 --> 00:19:30.820
Derek Peters: One question at a time.

189
00:19:31.040 --> 00:19:33.060
Derek Peters: Don't invent the answers.

190
00:19:33.650 --> 00:19:39.930
Derek Peters: And push back… You know, have it push back when you're vague, like, when you're vague.

191
00:19:40.540 --> 00:19:43.200
Derek Peters: That's a planner's prompt, not a wish.

192
00:19:43.380 --> 00:19:46.269
Derek Peters: And the card beside it is…

193
00:19:46.590 --> 00:19:49.749
Derek Peters: what it came back with. A PVD.

194
00:19:49.910 --> 00:19:56.879
Derek Peters: So… let's say, through that conversation. So, act as a product coach. Interview me one question at a time.

195
00:19:57.050 --> 00:19:58.360
Derek Peters: about cadence.

196
00:19:59.250 --> 00:20:08.889
Derek Peters: the user, the core problem, the metric that matters, and the three principles. Then draft a one-page PVD. Don't invent answers if I'm vague.

197
00:20:09.090 --> 00:20:10.940
Derek Peters: Push back until it's sharp.

198
00:20:12.300 --> 00:20:14.319
Derek Peters: And then have it interview you.

199
00:20:14.470 --> 00:20:19.480
Derek Peters: And go through, and then that way you can kind of get all the stuff that's in your head kind of out there.

200
00:20:19.690 --> 00:20:28.030
Derek Peters: And you also give it direction, and now it knows exactly what role it has to play, and in return, it's going to help you flesh out that vision.

201
00:20:28.810 --> 00:20:33.210
Derek Peters: I do that all the time, like, if I'm struggling to,

202
00:20:34.080 --> 00:20:38.850
Derek Peters: to kind of get what… what I'm thinking about. I… I… instead of just…

203
00:20:39.960 --> 00:20:47.719
Derek Peters: typing, and then have it respond, and then keep going back, I have it interview me. I give it a… I give it a title, and have it interview me.

204
00:20:51.860 --> 00:20:55.220
Derek Peters: It really depends, so this is for Mark.

205
00:20:55.400 --> 00:21:00.029
Derek Peters: I do like the kind of the interview flow, where, like.

206
00:21:00.280 --> 00:21:15.899
Derek Peters: we take it one bit at a time. The LLMs have gotten to a point now where you could probably do everything all at once and answer all those questions, but for the way that I process information, it's good for me to get it back, you know, kind of turn by turn.

207
00:21:16.240 --> 00:21:21.760
Derek Peters: And… answer, and then have it clarify, answer again.

208
00:21:21.940 --> 00:21:35.459
Derek Peters: But you could have it go to ask question user flow. My preference is turn by turn, just because I know how I work and operate, and that works best for me. So whatever works best for you, but that's a great question.

209
00:21:37.780 --> 00:21:41.669
Derek Peters: So let's talk about, turn a vision into a PRD.

210
00:21:43.640 --> 00:22:01.330
Derek Peters: So this is where you're… this is where you're defining and kind of fine-tuning. So the vision is a way… well, I'm sorry, the vision is the why, the PRD is the what. So anytime that you're developing, anytime that I'm thinking about what I'm going to do, I always want to…

211
00:22:01.620 --> 00:22:05.519
Derek Peters: I always want to ask myself, why am I doing this?

212
00:22:06.460 --> 00:22:11.200
Derek Peters: And then get into… what? What are we doing?

213
00:22:11.750 --> 00:22:13.559
Derek Peters: That satisfies the why.

214
00:22:13.950 --> 00:22:20.829
Derek Peters: You can even go back to, like, the user story logic. As a user, I want to do this, so that I can do that.

215
00:22:22.410 --> 00:22:27.099
Derek Peters: Because that way you have the who, The what and the why.

216
00:22:28.300 --> 00:22:35.369
Derek Peters: But… And, and you have one ver… you know, let's see…

217
00:22:36.100 --> 00:22:50.600
Derek Peters: You have the vision, which is the why, the PRD, which is the Y, and only for version 1, you have the PRD. So, the prompt on this slide, and I want you to… I want to call this out. Notice how much instruction I'm packing into my prompt.

218
00:22:50.930 --> 00:23:06.089
Derek Peters: Because that's the whole difference between a vague ask, that's the difference between vibe coding, and a useful one. That's really going to drill down, it's actually going to save you some time, too, from tokens, the clearer that you can be with the LLM and interact with it. The lot…

219
00:23:06.510 --> 00:23:11.570
Derek Peters: you're going to get a more richer response. So I'll point the model at the pvd.md.

220
00:23:12.290 --> 00:23:14.200
Derek Peters: As the single source of truth.

221
00:23:16.010 --> 00:23:20.349
Derek Peters: And… I name the exact sections that I want.

222
00:23:22.700 --> 00:23:29.159
Derek Peters: and then… well, and I… what I want to do is I name the… not only do I name the exact sections that I want back.

223
00:23:30.660 --> 00:23:35.810
Derek Peters: I have 3 to… you know, give me 3… 3 to 5 user stories.

224
00:23:35.960 --> 00:23:39.969
Derek Peters: Give me in scope versus out of scope, and measurable success.

225
00:23:40.410 --> 00:23:43.529
Derek Peters: Because that's what you need to do as a product planner.

226
00:23:46.680 --> 00:23:49.950
Derek Peters: I'll tell it to cut anything that's non-essential.

227
00:23:50.120 --> 00:23:51.990
Derek Peters: And park it under out of scope.

228
00:23:52.430 --> 00:23:57.719
Derek Peters: And I saw the comment in the… yes, that's… I always do that because…

229
00:23:58.030 --> 00:24:09.970
Derek Peters: Sometimes we tend to take on more than we can chew, and knowing clearly on the front end what's out of scope and what's in a scope can help as you develop. And then you can set some stretch goals.

230
00:24:12.780 --> 00:24:14.700
Derek Peters: And then I sequence with a roadmap.

231
00:24:15.450 --> 00:24:16.950
Derek Peters: So, this is…

232
00:24:17.060 --> 00:24:24.789
Derek Peters: This is what I refer to, you know, in the planning phase, this is what I like to call sequence. And you're turning that PRD into a roadmap.

233
00:24:25.130 --> 00:24:34.020
Derek Peters: It's… And people often get confused, like, this is not a wish list, this is, like, hard must-haves.

234
00:24:34.700 --> 00:24:40.549
Derek Peters: You might have nice-to-haves on the back end somewhere, kind of listed as your stretch goals.

235
00:24:40.880 --> 00:24:44.550
Derek Peters: But this… this is not a wish list sorted by what sounds fun.

236
00:24:44.810 --> 00:24:51.360
Derek Peters: And this is where a lot of people, when they're planning, they miss out. Like, sequence by risk and value.

237
00:24:52.220 --> 00:24:58.420
Derek Peters: And notice the prompt does the thinking for the model. So, sequence by risk times value.

238
00:24:58.830 --> 00:25:04.320
Derek Peters: Not by excitement, Because you're only in the infancy of this planning.

239
00:25:05.540 --> 00:25:09.879
Derek Peters: And put only the riskiest assumptions work in now?

240
00:25:09.940 --> 00:25:26.939
Derek Peters: and demand… you could demand, like, a one line of rationale per item if you need help making a decision on where to move forward. But here's the roadmapap that it gave back. And just so you all know, like, I went through this exercise, and…

241
00:25:27.510 --> 00:25:33.700
Derek Peters: condensed the output, but what you're seeing is the output. Now, next, later.

242
00:25:33.990 --> 00:25:34.820
Derek Peters: So…

243
00:25:35.650 --> 00:25:45.680
Derek Peters: And that's a really great tool to follow, because that way it keeps you on track, and it keeps you… keeps a nice sequence as you go into developing.

244
00:25:46.080 --> 00:25:49.730
Derek Peters: Yes, when I do this, they live in a GET repo.

245
00:25:50.060 --> 00:25:55.710
Derek Peters: And, they're usually README files, MD files, that I put into a docs folder.

246
00:25:56.620 --> 00:25:59.099
Derek Peters: But you could put them in,

247
00:25:59.270 --> 00:26:09.250
Derek Peters: Notion or whatever other tool that you would like to utilize, but I mostly keep them in my repo. That way, I can always have everything in one place.

248
00:26:15.380 --> 00:26:23.700
Derek Peters: The next question is, how do you communicate with the dev teams? The way that I communicate to the dev teams is, during stand-ups, or let's say there's

249
00:26:25.100 --> 00:26:35.419
Derek Peters: like, a once-a-week product, like, a planning session to, like, for plan… like, you have, like, a product planning session where you decide, like, what tasks are going to get worked out.

250
00:26:35.940 --> 00:26:44.229
Derek Peters: that is typically, like, an hour long, and then we follow up with, like, daily stand-ups. Just like a scaled, Agile or,

251
00:26:45.610 --> 00:26:53.759
Derek Peters: kind of a scrum framework, like, none of that really changes. What changes is how much you plan for your sprint in this new world of…

252
00:26:54.100 --> 00:26:55.290
Derek Peters: technology.

253
00:26:57.820 --> 00:27:11.769
Derek Peters: Yeah, they would appear into, into the, like, whatever repo is for, like, the product or the feature that you're working on. Typically we keep them in the… you would want to keep it in the repo themselves, like, of the product repo.

254
00:27:11.930 --> 00:27:16.740
Derek Peters: And then they can access those. Now, those were some great questions. Thank you so much.

255
00:27:18.910 --> 00:27:19.580
Derek Peters: Okay.

256
00:27:21.270 --> 00:27:28.849
Derek Peters: And the last planning move, and it… this is where it kind of compounds. You've got 3 documents now.

257
00:27:30.850 --> 00:27:33.470
Derek Peters: Don't paste them into the chat and lose them.

258
00:27:33.910 --> 00:27:43.929
Derek Peters: put them into a durable context file that AI reads every single time, And here's… Here's cadence.

259
00:27:45.150 --> 00:27:59.730
Derek Peters: This is the product that we're working on. So it points to the vision document, and points at version 1 of the requirements document. Because remember, this is… the reason why this is V1 is this is the first iteration of this product.

260
00:27:59.970 --> 00:28:06.229
Derek Peters: This is an… it's an ideation, that's where we started. Now we got something kind of fleshed out.

261
00:28:08.770 --> 00:28:11.349
Derek Peters: Yes, we're definitely going to get into that, Christian.

262
00:28:11.600 --> 00:28:19.659
Derek Peters: Not so much the architecture, because this is more product planning for tonight, but but yeah, we'll talk about prompting.

263
00:28:20.130 --> 00:28:21.560
Derek Peters: That that'd be next.

264
00:28:23.890 --> 00:28:25.050
Derek Peters: And…

265
00:28:25.910 --> 00:28:36.660
Derek Peters: So the PVD points to version 1 of the PRD, because that PRD is a living document, it's going to change as more requirements come about, and as you get… as you continue to progress your product.

266
00:28:40.440 --> 00:28:51.379
Derek Peters: and one rule that really matters for the… you'll also, like, let's say in that, PRD, as one thing, one thing that might live in there is, like.

267
00:28:52.260 --> 00:28:55.910
Derek Peters: We'll have the definition of done, and then,

268
00:28:56.380 --> 00:29:01.010
Derek Peters: maybe, like, whatever AI or whatever features that you're planning to, to integrate.

269
00:29:02.410 --> 00:29:09.639
Derek Peters: But the beauty of this is, I never have to re-explain what… Cadence is, again.

270
00:29:09.890 --> 00:29:14.600
Derek Peters: And every prompt inherits the vision and the scope automatically.

271
00:29:15.090 --> 00:29:26.740
Derek Peters: it… it gets treated like code, especially, and this is the beauty of having it in your repo, into a, like, a separate Docs version, or into a… I'm sorry, into a separate docs folder.

272
00:29:27.220 --> 00:29:32.280
Derek Peters: And you can version it, and then you can tell it, like, which version to go. Like, you could have multiple

273
00:29:32.700 --> 00:29:35.170
Derek Peters: Iterations of it as you continue.

274
00:29:35.330 --> 00:29:38.849
Derek Peters: But you can version it, you can review it, you can keep it current.

275
00:29:39.350 --> 00:29:42.579
Derek Peters: A stale context file can be dangerous.

276
00:29:43.070 --> 00:29:48.790
Derek Peters: So, make sure that it's always pointing to the right documentation.

277
00:29:49.020 --> 00:29:52.159
Derek Peters: Because the model will trust it completely.

278
00:29:52.980 --> 00:29:59.949
Derek Peters: And… You being kind of the captain of that ship, and kind of making sure that it's on track?

279
00:30:00.070 --> 00:30:03.160
Derek Peters: Definitely, I would recommend using version control.

280
00:30:03.390 --> 00:30:06.749
Derek Peters: But this is all V1, because this is a new idea.

281
00:30:12.150 --> 00:30:13.830
Derek Peters: So let's talk about prompting.

282
00:30:18.440 --> 00:30:21.290
Derek Peters: Yeah, you can convert anything to an ELD.

283
00:30:22.730 --> 00:30:33.840
Derek Peters: there was a question in here that I saw. Do you manage memory separate from PRD or PVD? I typically just use version control, so it knows which one to point to.

284
00:30:35.230 --> 00:30:40.220
Derek Peters: And then I'll archive. I don't know if you can elaborate on that question, Mark.

285
00:30:40.370 --> 00:30:42.130
Derek Peters: I'll be more than happy to answer it.

286
00:30:42.300 --> 00:30:48.689
Derek Peters: But let's get into prompting. So prompting, this is the part that…

287
00:30:48.930 --> 00:30:51.749
Derek Peters: I hope a lot of you get out of this.

288
00:30:52.720 --> 00:30:55.730
Derek Peters: But we wrote 3 documents before getting here.

289
00:30:56.070 --> 00:31:02.460
Derek Peters: Three of the five moves in the loop that we talked about, you know, going back to the methodology.

290
00:31:02.660 --> 00:31:08.800
Derek Peters: And now those documents pay off enormously.

291
00:31:09.370 --> 00:31:12.720
Derek Peters: So the mental model that I want you to adopt

292
00:31:14.060 --> 00:31:16.900
Derek Peters: You are not a user of the model.

293
00:31:18.540 --> 00:31:27.980
Derek Peters: You are the tech lead. You are a product manager of that model. So clear intent goes in, reviewable work comes out.

294
00:31:28.160 --> 00:31:31.709
Derek Peters: And you put your name on all of it. Everything in…

295
00:31:31.870 --> 00:31:39.579
Derek Peters: And this next section is about getting the signal in and keeping the loop tight enough that you never lose control of it.

296
00:31:42.270 --> 00:31:49.589
Derek Peters: Oh, yeah, yeah, you can do that. Sometimes you can also cache some of that, but yes. No, that's good.

297
00:31:53.060 --> 00:31:54.109
Derek Peters: Thank you, Mark.

298
00:31:54.500 --> 00:31:58.339
Derek Peters: Alright, so let's get into it here. So, a re… a weak prompt…

299
00:31:59.050 --> 00:32:15.649
Derek Peters: And we've all started here when AI first came out, and we're like, oh my gosh, what can we do with this? And we just, like, asked it all kinds of crazy questions, but it's… has evolved to a powerful tool that we all use today. But a weak prompt says, build

300
00:32:15.810 --> 00:32:19.029
Derek Peters: let's say, build the AI thing that makes tasks.

301
00:32:19.380 --> 00:32:22.090
Derek Peters: Or build me a task management system.

302
00:32:24.650 --> 00:32:29.670
Derek Peters: A strong prompt says three things in one… Breath.

303
00:32:30.730 --> 00:32:37.070
Derek Peters: what do you want? The constraints it has, you know, the constraints it has to respect.

304
00:32:37.410 --> 00:32:41.020
Derek Peters: And you'll know both, or you'll…

305
00:32:41.390 --> 00:32:44.599
Derek Peters: you'll both know it's done. Or…

306
00:32:44.920 --> 00:32:46.979
Derek Peters: How you'll… like, how do you know it's done?

307
00:32:47.220 --> 00:32:51.099
Derek Peters: It will tell you that. And the highest leverage habit

308
00:32:51.670 --> 00:32:54.969
Derek Peters: There is, ask for a plan before any code.

309
00:32:55.200 --> 00:32:59.040
Derek Peters: The anatomy is… On the slide here.

310
00:32:59.610 --> 00:33:03.519
Derek Peters: Intent, context, rules, and acceptance.

311
00:33:03.890 --> 00:33:06.590
Derek Peters: So, what and why in one sentence.

312
00:33:06.770 --> 00:33:09.419
Derek Peters: Files, stack, patterns to follow.

313
00:33:09.960 --> 00:33:15.069
Derek Peters: straight from the PRD, or a context.md, or a context.md.

314
00:33:15.440 --> 00:33:19.350
Derek Peters: and accept the PRD's own success metrics.

315
00:33:20.930 --> 00:33:26.950
Derek Peters: You make the model show you its plan first, That way, you understand

316
00:33:27.550 --> 00:33:33.160
Derek Peters: Clearly, what's going to happen next, and then you can help kind of direct it and guide it where to go.

317
00:33:36.340 --> 00:33:48.140
Derek Peters: I haven't used plan mode yet. I may give that a try, but I usually do, I usually put it into, like, a project folder, and then plan, and then transfer it into GitHub.

318
00:33:48.480 --> 00:33:55.960
Derek Peters: Using, Quad Desktop. That's my preferred method. I'm not saying that's the method to use. Everybody…

319
00:33:56.310 --> 00:33:59.879
Derek Peters: This kind of flows regardless what you want to have.

320
00:34:01.440 --> 00:34:06.320
Derek Peters: And I'm learning from you all just as much as I hope you're learning from me.

321
00:34:07.280 --> 00:34:08.810
Derek Peters: That's good.

322
00:34:09.449 --> 00:34:10.830
Derek Peters: forward slash goal.

323
00:34:10.960 --> 00:34:12.520
Derek Peters: In Codex, alright.

324
00:34:13.460 --> 00:34:19.149
Derek Peters: So, once the plan's approved, you have a model… well, you… you hand the model a map.

325
00:34:19.920 --> 00:34:30.429
Derek Peters: And the map literally… well, and the map is literally the prompt on the slide, so walk through it. First, I pointed at the docks.

326
00:34:30.730 --> 00:34:33.700
Derek Peters: Read the PRD and the context file.

327
00:34:34.150 --> 00:34:38.210
Derek Peters: because… Because the docs are the spec.

328
00:34:40.010 --> 00:34:42.900
Derek Peters: You're basically feeding it everything it needs.

329
00:34:43.250 --> 00:34:45.679
Derek Peters: And the exact files to touch.

330
00:34:46.400 --> 00:34:49.409
Derek Peters: And an existing pattern to follow.

331
00:34:50.120 --> 00:34:54.749
Derek Peters: That one line is worth a whole paragraph of style notes.

332
00:34:55.340 --> 00:34:58.500
Derek Peters: So the… the rules come straight from the PRD.

333
00:34:59.220 --> 00:35:07.929
Derek Peters: return, return type JSON, Set the priority and the urgency in the text. Never invent a due date.

334
00:35:08.810 --> 00:35:13.080
Derek Peters: And it's also a great failsafe for bad output.

335
00:35:13.620 --> 00:35:14.940
Derek Peters: then…

336
00:35:16.050 --> 00:35:25.539
Derek Peters: let's say, then done when is the PRD's own success metric. So, a prioritized edible list in under 5 seconds.

337
00:35:26.170 --> 00:35:27.220
Derek Peters: Plus…

338
00:35:27.590 --> 00:35:34.339
Derek Peters: the tests that have to pass, because that's another thing, when you don't properly plan, a lot of people don't realize

339
00:35:34.600 --> 00:35:42.650
Derek Peters: Oh, you mean I have to do testing? That's just as important as planning, like, once you get to the end, then you're ready to test and roll it out to real users?

340
00:35:44.900 --> 00:35:49.030
Derek Peters: So… Now you have everything that you need to go forward.

341
00:35:52.010 --> 00:35:56.969
Derek Peters: So I'm just feeding the model the documents we already wrote, the PRD is the prompt.

342
00:35:57.230 --> 00:36:04.939
Derek Peters: On the next slide, we'll watch what happens, let's say, when you skip all of this. And let's… let's talk about, like, a good prompt versus a bad prompt.

343
00:36:05.510 --> 00:36:09.029
Derek Peters: So this slide… Is that the problem?

344
00:36:09.140 --> 00:36:21.139
Derek Peters: you'll see there's a good prompt versus a bad one, and again, we've all been there, like, especially when AI started rolling out, and then it started getting a lot more richer, and then now it's a tool that we use every day.

345
00:36:24.000 --> 00:36:30.289
Derek Peters: So, on the left is the bad prompt. Add an AI feature that turns notes into tasks. Make it smart.

346
00:36:30.460 --> 00:36:42.790
Derek Peters: And here's what you actually get. You know, what you get is it splits text into breaks. It will deliver you something that does somewhat work, you get excited, but there's no real proof behind it.

347
00:36:44.200 --> 00:36:48.900
Derek Peters: And then on the right is an example of a good prompt.

348
00:36:49.680 --> 00:36:54.080
Derek Peters: Implement, extract, and prioritize PRDV1,

349
00:36:54.200 --> 00:37:05.609
Derek Peters: You can point it to it. You can, you don't even have… you don't have to give it its… the whole file path, unless you have multiple versions. Especially in the first prompt, you want to be clear on, like, where it pulls from.

350
00:37:08.210 --> 00:37:15.110
Derek Peters: And then… return type JSON title priority due, except brain dump.

351
00:37:15.280 --> 00:37:26.660
Derek Peters: edible list in less than 5 seconds, plan first, no code until I approve. That's important, because if you're running auto mode, or something that just kind of goes on its own.

352
00:37:26.890 --> 00:37:34.409
Derek Peters: You lose a little sense of control, even though it can do something really awesome and really cool, and you're like, oh, wow, this is… this is creating something.

353
00:37:34.540 --> 00:37:38.159
Derek Peters: Always ask it to approve.

354
00:37:39.200 --> 00:37:43.019
Derek Peters: Don't… don't just have it go and create a bunch of stuff.

355
00:37:45.240 --> 00:37:49.180
Derek Peters: So here's the rhythm that keeps it on the rails.

356
00:37:49.330 --> 00:37:53.340
Derek Peters: Ask, review, test, refund, and repeat.

357
00:37:54.920 --> 00:37:57.970
Derek Peters: Small differences over big rewrites.

358
00:37:58.230 --> 00:37:59.570
Derek Peters: every time.

359
00:37:59.680 --> 00:38:12.069
Derek Peters: Because a small diff is something you can actually verify. Read every line. I know it's tempting to skim, or skip, or just be like, yeah, get what I needed and go forward.

360
00:38:12.560 --> 00:38:16.129
Derek Peters: and… It looks confident, but you own this.

361
00:38:16.610 --> 00:38:27.029
Derek Peters: Not the model. So, like, if… now, if you're doing something on your own, just to help yourself out, that's fine, but if you're… if you're shipping something that users are going to use.

362
00:38:27.330 --> 00:38:28.280
Derek Peters: Read it.

363
00:38:28.950 --> 00:38:30.660
Derek Peters: Even if it looks confident.

364
00:38:31.690 --> 00:38:36.639
Derek Peters: If, you know, the product that we're working on.

365
00:38:36.950 --> 00:38:40.970
Derek Peters: If Cadence quietly buries, call someone back.

366
00:38:41.070 --> 00:38:44.229
Derek Peters: That's an important task at the bottom as low priority.

367
00:38:44.400 --> 00:38:52.659
Derek Peters: the AI is not the one who has to answer for it. If you miss a task, and you're creating, let's say, this task management system.

368
00:38:53.800 --> 00:39:01.970
Derek Peters: And it's… Not prioritizing right, or taking these tasks and doing what it's supposed to with them.

369
00:39:03.030 --> 00:39:06.979
Derek Peters: then that's… The person that develops its issue.

370
00:39:09.260 --> 00:39:10.320
Derek Peters: And…

371
00:39:11.610 --> 00:39:21.810
Derek Peters: let's say, the move that people that I see come through, that people resist most, is when the model gets stuck and starts circling. Don't argue with it.

372
00:39:22.090 --> 00:39:26.990
Derek Peters: Just reset the context, open with a clean chat, start fresh.

373
00:39:28.390 --> 00:39:42.590
Derek Peters: And I know that, I mean, as good as these models are that are out here today, they can still get confused, they can still hallucinate. Arguing with a confused model just trains it to be confidently wrong.

374
00:39:43.950 --> 00:39:49.860
Derek Peters: at you. So, and… If you're using Clyde, sometimes it's very confident.

375
00:39:50.220 --> 00:39:51.859
Derek Peters: And you just gotta restart.

376
00:39:52.060 --> 00:39:54.139
Derek Peters: And just kind of reset the context.

377
00:39:59.420 --> 00:40:07.489
Derek Peters: Yeah, I do manually reset the context window, or I'll, I'll have, like…

378
00:40:08.500 --> 00:40:15.939
Derek Peters: a golden prompt that I can use to kind of start a new chat and put it in there. So, like, I'll save some prompts in the back end.

379
00:40:16.890 --> 00:40:28.199
Derek Peters: That's true, codecs compacting, that is such a great feature. Absolutely. And just have it compact the con… and do the context resetting itself, that is pretty solid.

380
00:40:29.500 --> 00:40:30.190
Derek Peters: What's happening?

381
00:40:30.890 --> 00:40:32.340
Derek Peters: Good feedback.

382
00:40:33.120 --> 00:40:40.369
Derek Peters: So, before we ship, Let me name 4 ways this all falls apart.

383
00:40:42.140 --> 00:40:49.860
Derek Peters: Because over a long session, You know, you went through all of this… You'll fill the form.

384
00:40:52.940 --> 00:40:57.779
Derek Peters: So… One, a 600 line diff.

385
00:40:58.640 --> 00:41:00.240
Derek Peters: You can't review.

386
00:41:00.850 --> 00:41:05.290
Derek Peters: Well, one, because it's 600 lines, so you don't, you just trust it.

387
00:41:06.180 --> 00:41:09.949
Derek Peters: Two, confident code that reads beautifully, and

388
00:41:10.640 --> 00:41:15.219
Derek Peters: It's quietly wrong. So you still need to go through it and just…

389
00:41:15.380 --> 00:41:20.649
Derek Peters: Check. Testing, too. Make sure you embed testing into your planning.

390
00:41:21.010 --> 00:41:28.589
Derek Peters: Three, arguing with the stuck model with the context rots into mush. You're just kind of going downhill, you're not going to win that battle.

391
00:41:29.790 --> 00:41:37.060
Derek Peters: Looks like you all have ran into this at some point with the context window and having to reset it, which is fantastic.

392
00:41:38.460 --> 00:41:40.930
Derek Peters: But keep in mind, the model didn't fail.

393
00:41:42.310 --> 00:41:43.739
Derek Peters: The workflow did.

394
00:41:45.290 --> 00:41:48.189
Derek Peters: And the fix is always the same three things.

395
00:41:48.660 --> 00:41:52.930
Derek Peters: smaller ask, Read the changes.

396
00:41:54.060 --> 00:42:00.630
Derek Peters: And implement or incorporate some sort of acceptance test as you're going through this.

397
00:42:02.080 --> 00:42:10.140
Derek Peters: when you're… when your AI sessions fall apart, And it will come back…

398
00:42:10.730 --> 00:42:17.130
Derek Peters: And just remember that, like, just kind of reset the context, remind it what it's doing, remind…

399
00:42:17.440 --> 00:42:25.120
Derek Peters: remind… yeah, reminded what it's doing, and just kind of move forward. Don't get aggravated, as aggravating as it can be, but

400
00:42:26.060 --> 00:42:28.879
Derek Peters: And keep your prompts clean and informative.

401
00:42:29.460 --> 00:42:31.019
Derek Peters: Yes, don't let it compact.

402
00:42:31.940 --> 00:42:33.540
Derek Peters: So let's talk about shipping.

403
00:42:34.200 --> 00:42:39.350
Derek Peters: So this is where AI-built products quietly… Can, if they're…

404
00:42:40.400 --> 00:42:44.669
Derek Peters: Vibe-coded, or no thoughts behind it, this is where they go to die.

405
00:42:45.290 --> 00:42:51.099
Derek Peters: It might be a beautiful demo on someone's laptop, never reaches a single user.

406
00:42:51.450 --> 00:42:56.390
Derek Peters: Because shipping started to feel scary, started to feel a little nervous.

407
00:43:01.730 --> 00:43:03.500
Derek Peters: So we're going to…

408
00:43:04.430 --> 00:43:20.840
Derek Peters: my goal is not to make shipping sound boring, but my goal is, you know, let's start small, let's flag it, let's measure it. The entire goal of shipping, and if we go back to, like, let's say Cadence, as an example, or task management software.

409
00:43:21.360 --> 00:43:33.689
Derek Peters: you know, the first slice in front of real people without anyone in the building holding their breath. Like, you want to make it to where, like, hey, I want to demo this to a user, or I'm going to go in.

410
00:43:33.840 --> 00:43:41.279
Derek Peters: Shipping is… this proof of concept, getting feedback, getting real user feedback. So let's talk about the ship

411
00:43:41.640 --> 00:43:43.420
Derek Peters: Ship behind a flag.

412
00:43:43.830 --> 00:43:47.960
Derek Peters: So… The trick to shipping fast?

413
00:43:48.600 --> 00:43:52.909
Derek Peters: Is not shipping fast, fast, fast, fast.

414
00:43:53.060 --> 00:44:02.839
Derek Peters: And what I mean by that is, just… It's okay to submit, but… Going too fast?

415
00:44:03.240 --> 00:44:10.939
Derek Peters: can sometimes cause problems. But the secret to shipping fast is shipping small, and shipping often.

416
00:44:11.290 --> 00:44:18.550
Derek Peters: So… with feature flags, And with feature flags.

417
00:44:18.680 --> 00:44:21.760
Derek Peters: That's what makes it… makes it safe.

418
00:44:23.270 --> 00:44:26.890
Derek Peters: You have those small little iterations that you can ship out.

419
00:44:27.360 --> 00:44:32.639
Derek Peters: But you're shipping in small chunks. You're shipping often, and you're testing those.

420
00:44:34.560 --> 00:44:39.780
Derek Peters: So let me say what shit behind the flag actually means. Well, what it means to me.

421
00:44:39.950 --> 00:44:45.469
Derek Peters: Because, to me, it's the whole foundation of the slide that you're seeing right now.

422
00:44:45.650 --> 00:44:50.269
Derek Peters: A feature flag is just an on-off switch wrapped around your new code.

423
00:44:50.730 --> 00:44:55.519
Derek Peters: If the flag is on, Run a new extractor. If it's off, don't.

424
00:44:55.690 --> 00:45:01.159
Derek Peters: You deploy the code to production with that switch turned off.

425
00:45:01.470 --> 00:45:04.400
Derek Peters: It lives… well, it lives on the server.

426
00:45:04.640 --> 00:45:07.320
Derek Peters: But it's invisible to users until you decide to flip it.

427
00:45:07.860 --> 00:45:14.419
Derek Peters: So… Shipping and turning on becomes two separate reversible steps.

428
00:45:14.850 --> 00:45:17.600
Derek Peters: And that's what takes the fear

429
00:45:17.750 --> 00:45:20.450
Derek Peters: Yeah, it takes the fear out of releasing.

430
00:45:20.610 --> 00:45:27.380
Derek Peters: So, you merged unfinished work, Because it's switched off in production.

431
00:45:28.130 --> 00:45:30.969
Derek Peters: That keeps Maine reasonable at all times.

432
00:45:31.310 --> 00:45:37.810
Derek Peters: Which means you are never trapped in, let's say, a 6-week branch that fights you on every rebase.

433
00:45:38.490 --> 00:45:42.699
Derek Peters: And if you've ran into merge conflicts, that can be aggravating.

434
00:45:43.990 --> 00:45:50.009
Derek Peters: Then, you roll out gradually. Internal users first, then a small percentage.

435
00:45:50.140 --> 00:45:59.200
Derek Peters: And if you're working, let's say, on a solo product, grab some folks that can give you some feedback on that, and just do some user testing as well.

436
00:45:59.680 --> 00:46:04.960
Derek Peters: But internal users first, Then a small percentage, watching the…

437
00:46:05.140 --> 00:46:07.119
Derek Peters: Keep an eye on it the whole way.

438
00:46:07.880 --> 00:46:10.269
Derek Peters: Then, you roll it out to everyone.

439
00:46:11.000 --> 00:46:14.189
Derek Peters: If, let's say, the extractor returns something weird.

440
00:46:14.300 --> 00:46:20.820
Derek Peters: You flip a, you know, let's… yeah, return something weird, you flip a switch, You don't need…

441
00:46:21.360 --> 00:46:25.049
Derek Peters: you don't need to do, like, an emergency 2 a.m. deployment.

442
00:46:25.770 --> 00:46:34.200
Derek Peters: the prompts… And the prompts on the sl… got tongue twisted there. And the prompts on the slide?

443
00:46:34.340 --> 00:46:48.569
Derek Peters: And it… you'll see it right there, like, wrap the extractor in a flag cadence, extract V1 to fall off with fallback to manual entry, roll out to internal 5%, 100% gated error rate less than 1%.

444
00:46:50.200 --> 00:46:57.759
Derek Peters: So… When you're shipping features, It's always good to ship small.

445
00:46:58.110 --> 00:47:04.710
Derek Peters: Shift often, and… Have a way to kind of easily cut them on and off for slow rollouts.

446
00:47:08.040 --> 00:47:10.120
Derek Peters: So here's a habit.

447
00:47:12.170 --> 00:47:14.580
Derek Peters: That most people miss completely.

448
00:47:15.220 --> 00:47:17.800
Derek Peters: They use AI to write code.

449
00:47:18.200 --> 00:47:21.160
Derek Peters: And then go back to reviewing all of it by hand.

450
00:47:21.280 --> 00:47:27.310
Derek Peters: That's complete… that's completely backwards, because I don't know if you've been in that situation, that's a lot of work.

451
00:47:27.970 --> 00:47:31.249
Derek Peters: Generate test alongside the feature.

452
00:47:31.400 --> 00:47:38.140
Derek Peters: not as an afterthought. If you have a feature, That you're planning to implement?

453
00:47:39.070 --> 00:47:41.279
Derek Peters: Work on that feature.

454
00:47:42.310 --> 00:47:43.980
Derek Peters: Test it.

455
00:47:45.350 --> 00:47:51.369
Derek Peters: If it's for users that, you know, if it's for real users that are going to use it, get some user feedback.

456
00:47:51.550 --> 00:47:53.019
Derek Peters: And then roll it out.

457
00:47:53.550 --> 00:47:55.700
Derek Peters: And test in increments.

458
00:47:57.980 --> 00:48:05.580
Derek Peters: So, the pop line on the slide runs the test. So, type checks, and it has an AI reviewer to read off the

459
00:48:06.180 --> 00:48:09.499
Derek Peters: Read off for what humans tend to skim past.

460
00:48:09.740 --> 00:48:12.570
Derek Peters: Leap Secrets, missing test.

461
00:48:12.970 --> 00:48:17.830
Derek Peters: broken conventions, and that's another thing that I want to call out, too. Like, the number…

462
00:48:19.470 --> 00:48:26.410
Derek Peters: I'm not saying, like, review it line by line, but if you're working on those small features, you want to…

463
00:48:26.910 --> 00:48:32.059
Derek Peters: Yeah, it's impossible to read, you know, read it line by line, but what I'm saying is.

464
00:48:33.440 --> 00:48:37.320
Derek Peters: Don't wait till you have… 6 features.

465
00:48:37.470 --> 00:48:41.359
Derek Peters: And thousands of lines of code in different files

466
00:48:42.420 --> 00:48:46.420
Derek Peters: to, to roll something out. Roll those features out in increments.

467
00:48:49.830 --> 00:48:56.249
Derek Peters: So… And make sure that it's testing, you know, generate tests with

468
00:48:56.440 --> 00:48:58.940
Derek Peters: With the feature, the happy path, empty dunk.

469
00:48:59.260 --> 00:49:12.999
Derek Peters: and off-schema response that fail, or that must fail safe, not crash, make the required CIG on every PR, and have it run through and just test your code. Have it read through your code, and do the testing.

470
00:49:17.170 --> 00:49:22.630
Derek Peters: So, shipping isn't the finish line. It's where you finally get real information.

471
00:49:23.910 --> 00:49:26.720
Derek Peters: So, instrument the feature before you announce it.

472
00:49:26.990 --> 00:49:29.709
Derek Peters: A dark lunch teaches nothing.

473
00:49:30.320 --> 00:49:38.020
Derek Peters: So… Back to our product, so if we take a look at Cadence first week, on the slide.

474
00:49:38.260 --> 00:49:42.559
Derek Peters: 61% of people pasted a brain dump.

475
00:49:43.570 --> 00:49:50.490
Derek Peters: Beating target, an average of, let's say, we got 61% pace of… Brain dump.

476
00:49:50.820 --> 00:49:57.450
Derek Peters: Target was 50%. Organized, so 6.4 tasks extracted per session.

477
00:49:57.880 --> 00:50:04.940
Derek Peters: 34% tweak a priority, and then next, you know, why is this priority? And then there's an explainer in there.

478
00:50:14.320 --> 00:50:20.890
Derek Peters: That's… yeah, that does harden the build. Each component should be a sub-dependent on itself, absolutely.

479
00:50:22.450 --> 00:50:23.120
Derek Peters: Do you think?

480
00:50:24.080 --> 00:50:27.680
Derek Peters: So let's bring this… To kind of full circle.

481
00:50:29.070 --> 00:50:37.529
Derek Peters: So, what we went over was writing a vision, a requirement document, and a roadmap before you even get into coding. Those are…

482
00:50:39.500 --> 00:50:44.590
Derek Peters: The three most important documents… to me.

483
00:50:44.860 --> 00:50:50.250
Derek Peters: That can be created and generated before you get into developing a product.

484
00:50:54.390 --> 00:51:01.540
Derek Peters: And… and those are, like, those are the spec the AI builds from. So… and then we talked about prompting.

485
00:51:02.160 --> 00:51:04.649
Derek Peters: We talked about, you know.

486
00:51:05.860 --> 00:51:12.370
Derek Peters: We drove the build straight out of the requirements document, planned first, we tightened the loops.

487
00:51:12.800 --> 00:51:15.669
Derek Peters: We had it go through each line.

488
00:51:16.160 --> 00:51:20.090
Derek Peters: We talked about, you know, testing, And…

489
00:51:20.560 --> 00:51:32.940
Derek Peters: as long as you keep those in mind as you're progressing to build, that's what separates you from, let's say, somebody that's just 5 coding, creating things, which is… is still great, but…

490
00:51:34.530 --> 00:51:37.539
Derek Peters: Is there proof in the real world? Probably not.

491
00:51:38.070 --> 00:51:43.640
Derek Peters: the better you can plan the what and the why and the who.

492
00:51:44.030 --> 00:51:46.320
Derek Peters: The better your product's gonna turn out.

493
00:51:47.460 --> 00:51:55.330
Derek Peters: And the quality of what you ship is decided long before and long after the model types a single character, a single line of code.

494
00:51:55.850 --> 00:51:59.409
Derek Peters: So, before we open it up, Two quick things.

495
00:52:05.040 --> 00:52:09.150
Derek Peters: What are some things, you know, if you're working on some things, and…

496
00:52:09.440 --> 00:52:18.680
Derek Peters: you've been toying around? It sounds like we have a lot of experienced individuals on the call, too. Like, it'd be interesting, like, what are you working on? That… that would be fun.

497
00:52:29.200 --> 00:52:30.170
Derek Peters: Okay.

498
00:52:32.640 --> 00:52:34.189
Derek Peters: I love the interaction.

499
00:52:34.490 --> 00:52:35.760
Derek Peters: You all are awesome.

500
00:52:36.800 --> 00:52:37.480
Derek Peters: Okay.

501
00:52:38.680 --> 00:52:44.339
Derek Peters: So, before I make, before we go on to the next slide, here's…

502
00:52:45.400 --> 00:52:47.279
Derek Peters: what I actually want you to do.

503
00:52:48.600 --> 00:52:51.839
Derek Peters: Because… because of talk, you don't…

504
00:52:53.100 --> 00:52:59.850
Derek Peters: Let's see, because you don't act on entertainment, you know, 3 moves, 3 things that you can do.

505
00:53:00.030 --> 00:53:06.150
Derek Peters: to kind of help you is write a one-page PVD on the next project or next idea that you have.

506
00:53:06.430 --> 00:53:10.030
Derek Peters: Turn it into a requirements document, scope it out.

507
00:53:10.470 --> 00:53:15.110
Derek Peters: Write out, you know, what's in scope, what's out of scope.

508
00:53:15.640 --> 00:53:35.520
Derek Peters: you know, that keeps it shippable, it keeps it simple, it also makes it easy to develop a roadmap that you can actually deploy, and then make an AI plan first. Like, work with the LLM to make a plan and a roadmap for you. So, before you write a single line of code, have it propose a plan, and then you just approve it.

509
00:53:35.650 --> 00:53:39.550
Derek Peters: No new tools, it sounds like it's everything that everybody's using.

510
00:53:40.270 --> 00:53:42.560
Derek Peters: But, but that's it.

511
00:53:44.680 --> 00:53:50.260
Derek Peters: And before we get into any Q&A that's on there,

512
00:53:51.010 --> 00:53:59.510
Derek Peters: We… we do have our applications open. If there's anybody that's interested in joining us here at Gauntlet, would love to invite you to apply.

513
00:53:59.780 --> 00:54:04.109
Derek Peters: And, I also want to say thank you so much for your time.

514
00:54:04.330 --> 00:54:05.399
Derek Peters: as well.

515
00:54:05.920 --> 00:54:11.440
Derek Peters: And, we'll also be sharing that information with you as well. Oh, thank you, Sunny. There it is.

516
00:54:14.310 --> 00:54:15.530
Derek Peters: So that's it.

517
00:54:15.810 --> 00:54:22.820
Derek Peters: We kind of went through one product example from this ideation to…

518
00:54:23.250 --> 00:54:26.689
Derek Peters: To shipping the, you know, shipping the features?

519
00:54:27.110 --> 00:54:33.689
Derek Peters: And kept it into an AI-first planning methodology, so AI-first product planning.

520
00:54:34.170 --> 00:54:45.599
Derek Peters: So, same loop, real documents, real prompts, and every prompt and every result that you saw here was actually from my iteration going through my planning process.

521
00:54:46.200 --> 00:54:53.490
Derek Peters: And I will tell you that it does… Make a huge difference.

522
00:54:54.350 --> 00:54:56.870
Derek Peters: The more planning you can do up front.

523
00:54:57.550 --> 00:55:04.480
Derek Peters: Gives you a richer product that… Users will actually use.

524
00:55:04.980 --> 00:55:10.160
Derek Peters: So, the floor is wide open. We got, we got a few minutes. If anybody does have any questions.

525
00:55:11.700 --> 00:55:14.100
Derek Peters: Be more than happy to answer that for you.

526
00:55:17.340 --> 00:55:31.929
Derek Peters: So, Mark, you asked, you know, where does the UI design fit into this? It fits in… it fits into the very beginning of, because, like, whenever you're planning, like, your… let's say your product vision.

527
00:55:32.530 --> 00:55:52.500
Derek Peters: and your product requirements, like the PRD, you could either do it one of two ways. You could have, like, a requirement document for, like, your UI, UX piece of it, and have it, like, two separate documents. I do like to fold it up into the same PRD, that way I'm not having to point, to two documents.

528
00:55:52.890 --> 00:56:01.929
Derek Peters: or two different versions of the documents, because it can sometimes confuse the LLM, but you can fold that into your requirements document. That's a good question.

529
00:56:02.250 --> 00:56:08.040
Derek Peters: So, UI design actually follows the same set. You have a vision, you have your requirements.

530
00:56:08.180 --> 00:56:11.019
Derek Peters: You know, the what, The why, the who.

531
00:56:11.350 --> 00:56:16.850
Derek Peters: And… It's just… it becomes part of your requirements.

532
00:56:23.020 --> 00:56:29.460
Derek Peters: So, Josh, you said, is there a benefit to writing out code by hand anymore when AI can just write it for me?

533
00:56:29.970 --> 00:56:36.720
Derek Peters: I do think there is a huge benefit in having that background and knowledge, because

534
00:56:37.790 --> 00:56:49.210
Derek Peters: even though, like, let's say you have AI write it, you got the advantage of knowing, like, if you have an issue or, like, a bug or something, knowing how to read and write code is still an important skill today.

535
00:56:49.850 --> 00:56:56.779
Derek Peters: Because that way, you don't have to have it completely wipe something. If you…

536
00:56:56.960 --> 00:57:01.400
Derek Peters: you know, hey, I got this bug, find it, and then you realize that it's, like.

537
00:57:01.760 --> 00:57:17.110
Derek Peters: it rewrote 3 different files, but if, you know, still knowing how to write code, or at least jump-starting, you can even jumpstart writing your… getting your setup and start coding yourself, and then have the AI kind of be, like, a coding assistant to help you out. You don't have to have it do it

538
00:57:17.310 --> 00:57:21.969
Derek Peters: for you, it does kind of defeat the personal, like, you know, AI-first development.

539
00:57:22.180 --> 00:57:28.140
Derek Peters: But… Knowing how to write code yourself can also help save time in debugging.

540
00:57:34.510 --> 00:57:39.810
Derek Peters: Renee, what shipping flag tools are you using? So…

541
00:57:39.970 --> 00:57:56.729
Derek Peters: I put that into… kind of build it into my roadmap. I like to use Claude Code to… so I have it tied into a roadmap, and then I have each feature kind of incremented and cut out, but I like to keep those into my docs folder, into a README file, and then,

542
00:57:57.040 --> 00:58:01.740
Derek Peters: point the LLM to it, and then that way we know exactly what, shipping

543
00:58:02.750 --> 00:58:09.879
Derek Peters: shipping flag, or what feature gets shipped, and which ones to flag. So I use Claude Code for that.

544
00:58:14.190 --> 00:58:24.470
Derek Peters: And I may have missed Abby's question, so Carl, thanks for bringing it up right there. So, as a PM, I find one of the biggest challenges is slicing AI features into user stories.

545
00:58:24.920 --> 00:58:29.380
Derek Peters: That are both technically feasible and valuable enough to gain support.

546
00:58:31.090 --> 00:58:34.570
Derek Peters: In a larger enterprise where AI adoption tends to be slower.

547
00:58:38.770 --> 00:58:50.179
Derek Peters: close the… let's see, I'm just… I'm reading through this question here. So, how do we close the gap between PM's Quick Planning and Development and Classic Engineering? I'd be interested in what tools you have today

548
00:58:51.010 --> 00:58:52.690
Derek Peters: On that.

549
00:58:55.410 --> 00:58:59.450
Derek Peters: And there's so many… or, like, what AI features…

550
00:58:59.590 --> 00:59:01.970
Derek Peters: It doesn't have to be, like, what you're planning.

551
00:59:03.660 --> 00:59:08.130
Derek Peters: I have not tested Fable against Opus yet.

552
00:59:08.610 --> 00:59:12.009
Derek Peters: I just got into it right before this call.

553
00:59:12.280 --> 00:59:17.639
Derek Peters: I, I can't… bring myself to use Fable just yet. I'm sure it's fantastic.

554
00:59:18.170 --> 00:59:23.259
Derek Peters: but… burning through those tokens when Opus is

555
00:59:23.770 --> 00:59:30.600
Derek Peters: In most cases, Sonnet's good for what I need to use it for. Opus, I use for, like, more complex tasks.

556
00:59:30.850 --> 00:59:36.270
Derek Peters: But for day-to-day, Sonnet's pretty good. I haven't even opened the door of Fable.

557
00:59:41.710 --> 00:59:45.819
Derek Peters: Well, there's the answer. Josh ran out in an hour, so…

558
00:59:51.700 --> 01:00:09.550
Derek Peters: Now, one thing I do like about Opus, though, is, like, whenever you don't… like, you have a vision and you want it to interview, I have found that it does do a really good job in interviewing you with some half-baked information and helping you flesh out some ideas into a true vision. It does a really good job with that.

559
01:00:11.920 --> 01:00:15.400
Derek Peters: Oh, AI tools to develop features, okay.

560
01:00:17.290 --> 01:00:23.950
Derek Peters: Yes, that is rough in an enterprise environment, because there's a lot of security protocols,

561
01:00:25.520 --> 01:00:29.340
Derek Peters: in place, and I've ran into that before, as well.

562
01:00:30.050 --> 01:00:37.250
Derek Peters: and I think from there, it's just more about…

563
01:00:37.900 --> 01:00:50.569
Derek Peters: selling the benefit of what those tools or features can do, but if you're going to do that, like, let's say Carl and, I think it was Abby, there's a lot of great chats here, but…

564
01:00:50.680 --> 01:00:54.580
Derek Peters: I would come up with, like, a use case, and…

565
01:00:54.810 --> 01:00:59.120
Derek Peters: And some proof, like, what you can actually do with it, and how… and…

566
01:01:00.020 --> 01:01:18.880
Derek Peters: really sell the value of the return on investment that those AI features are going to help you as, let's say, a PM, and work with the team of developers. So an ROI, especially in an enterprise environment, is going to be super important, but the most important piece is the security piece, like.

567
01:01:18.950 --> 01:01:27.919
Derek Peters: what guardrails are you going to put around it to make sure that that enterprise information stays safe? Because that's what those corporations care about.

568
01:01:28.130 --> 01:01:34.839
Derek Peters: So, like, if you just come up with a good story and a good use case and show, like, what the return on investment is for the organization.

569
01:01:35.100 --> 01:01:41.269
Derek Peters: I've had a lot of good luck, getting ears to listen, I should say.

570
01:01:43.630 --> 01:01:44.939
Derek Peters: Does that help, Carl?

571
01:01:51.090 --> 01:01:53.180
Derek Peters: He turns into a house of cards.

572
01:01:57.300 --> 01:01:59.269
Derek Peters: All right.

573
01:02:00.330 --> 01:02:08.600
Derek Peters: Well, it does look like we're 2 minutes after. A lot of great conversation, a lot of great interactions, loved every minute of it.

574
01:02:08.970 --> 01:02:24.150
Derek Peters: Just wanted to say thank you all for allowing me to host you this evening for Gauntlets Night School, and really appreciate you coming. I hope you guys have a wonderful night, morning, and evening, depending on where you're at in the world. It's been great. Thank you all.

575
01:02:39.950 --> 01:02:43.469
Tom Babb: You can just log out, Derek, or you can just end the call.

576
01:02:43.700 --> 01:02:48.070
Derek Peters: Okay. Oh, no, I can't end it. It just lets me to leave, so you and Sonny were gonna…

577
01:02:48.070 --> 01:02:50.060
Tom Babb: Yeah, you got it, you got it. We're good, thank you.