WEBVTT 00:00.120 --> 00:01.720 So there's one more thing to mention. 00:01.720 --> 00:04.720 One more way in which it's just slightly more complicated. 00:05.240 --> 00:13.240 And that is that Landgraf itself is, in fact three different things, as I alluded to this already. 00:13.240 --> 00:20.560 But Landgraf is in fact Landgraf itself, which is also thought of as the Landgraf framework. 00:21.320 --> 00:28.200 It's a user interface tool called Landgraf Studio, which is one of these kinds of ways that you can 00:28.240 --> 00:29.440 hook things up visually. 00:29.440 --> 00:31.520 So like a visual builder. 00:32.160 --> 00:37.600 And it's something called Landgraf platform, which is the thing that they're sort of promoting on their 00:37.600 --> 00:46.280 website, as if that is Landgraf, uh, Landgraf platform, which is their, um, hosted solution for 00:46.280 --> 00:50.160 deploying and running your agents at scale. 00:50.680 --> 00:58.120 So, Landgraf, which is the framework which is analogous to Cray's framework, Landgraf studio, I 00:58.120 --> 01:00.560 think in Korea it was also called Korea Studio, right? 01:00.880 --> 01:01.440 I think so. 01:01.680 --> 01:06.600 And then Landgraf platform is analogous to CRI enterprise. 01:06.960 --> 01:09.550 And so these are the three offerings. 01:09.750 --> 01:14.590 And I mean, I think to me, I would make the same point that I made with crew, which is that what 01:14.750 --> 01:21.070 what perhaps part of what's going on here is, of course, Landgraf is looking for ways to commercialize 01:21.070 --> 01:22.910 and to monetize their offerings. 01:22.950 --> 01:30.270 And so looking for the enterprise play, the Landgraf platform is about deploying and running your Landgraf 01:30.270 --> 01:35.950 your graphs in their environment, uh, taking advantage of the fact that they've they've got hooks 01:35.950 --> 01:37.470 into everything that they've built. 01:37.470 --> 01:42.390 So I'm sure that if you have built all of your software using Landgraf, then it will be very convenient 01:42.390 --> 01:43.990 to use Landgraf platform. 01:43.990 --> 01:48.910 But I suspect that's what's going on, and I imagine that's why it is sort of promoted and framed so 01:48.910 --> 01:52.910 heavily on their site, as if Landgraf and Landgraf platform are the same thing. 01:53.030 --> 01:59.070 Uh, it's probably because this is the core commercial idea, uh, for for Lange chain and building 01:59.070 --> 01:59.630 this out. 01:59.950 --> 02:05.910 But we are going to be focused, of course, uh, as you can imagine on Landgraf, the framework, that 02:05.910 --> 02:07.550 part of it, that's what we're building to. 02:07.590 --> 02:09.070 We're going to be building our own. 02:09.270 --> 02:13.030 And we'll also be using Lange Smith as well to see what's going on. 02:13.150 --> 02:15.310 Uh, that's going to be interesting for us too. 02:15.350 --> 02:21.250 Now, at this point, I also wanted to do something I wanted to show you the blog post about building 02:21.290 --> 02:26.250 effective agents on Anthropic's website that I mentioned way back in week one, when we were looking 02:26.250 --> 02:27.330 at design patterns. 02:27.330 --> 02:33.010 But I think it's interesting and relevant now because it does show Anthropic's sort of positioning on 02:33.010 --> 02:37.090 this, and it's useful for you to keep this in mind when you think about the difference between what 02:37.090 --> 02:41.010 we're going to do today and what we're going to be doing in week six. 02:41.250 --> 02:48.570 So if we switch to building effective agents, then this this is the blog post you'll find on their 02:48.570 --> 02:49.250 website. 02:49.250 --> 02:51.050 And it's a brilliantly written post. 02:51.050 --> 02:56.850 I've got to tell you, it's really clear it contains within it the different design patterns that I 02:56.850 --> 02:57.290 mentioned. 02:57.330 --> 03:03.130 I hope I think I mentioned that I took the anthropic design patterns right from here. 03:03.170 --> 03:05.890 I thought they're so clear, so, so well explained. 03:06.130 --> 03:13.170 Um, and in addition to that, it has some discussion about using abstraction layers. 03:13.170 --> 03:15.210 And that's really what I wanted to show you here. 03:15.490 --> 03:22.250 So the, the, uh, um, it's says when and how to use frameworks. 03:22.250 --> 03:28.320 So they say there are many frameworks that make agentic systems easier to implement, including Landgraaf 03:28.320 --> 03:29.720 from Lange chain. 03:29.880 --> 03:35.440 They mention a few others that are not actually so popular, and not as far as I know, although I know 03:35.480 --> 03:36.320 of vellum the business. 03:36.320 --> 03:44.080 But but I don't think that that quite is as big as things like Cru and Autogen and now OpenAI agents 03:44.080 --> 03:44.600 SDK. 03:45.840 --> 03:48.000 But anyway, this is the point I wanted to make. 03:48.040 --> 03:53.160 These frameworks make it easy to get started by simplifying standard, low level tasks like calling 03:53.200 --> 03:57.960 llms, defining and parsing tools, and chaining calls together. 03:58.000 --> 04:03.640 However, they often create extra layers of abstraction and that can obscure the underlying prompts 04:03.640 --> 04:05.880 and responses, making them harder to debug. 04:06.280 --> 04:11.720 They can also make it tempting to add complexity when a simpler setup would suffice. 04:12.360 --> 04:14.320 You can see what Anthropic's getting at. 04:14.320 --> 04:17.400 You can imagine from their point of view, they've got an API. 04:17.720 --> 04:20.320 They've got something that is relatively simple. 04:20.360 --> 04:22.880 Memory can be handled with JSON objects. 04:23.280 --> 04:26.280 Llms can be hooked up simply by calling multiple times. 04:26.680 --> 04:33.600 And so for them, this this idea of building all of this abstraction around it, taking you further 04:33.660 --> 04:40.380 from actually working with the LM itself isn't necessarily something that resonates strongly. 04:41.140 --> 04:46.780 And so they conclude with we suggest that developers start by using LM APIs directly. 04:46.940 --> 04:49.660 Many patterns can be implemented in a few lines of code. 04:49.820 --> 04:53.900 If you do use a framework, ensure you understand the underlying code. 04:54.060 --> 04:58.220 Incorrect assumptions about what's under the hood are a common source of customer error. 04:58.460 --> 05:02.220 And by customer they mean us customers of anthropic. 05:02.900 --> 05:03.700 So there you go. 05:03.740 --> 05:05.140 I found this really interesting. 05:05.140 --> 05:08.780 It's obviously very clearly written and something that I'm very passionate about. 05:08.780 --> 05:10.940 It is an alternative, a different school of thought. 05:11.060 --> 05:20.460 It is, of course, somewhat, uh, not not in sync with the graph, uh, philosophy thesis of building 05:20.460 --> 05:21.660 this kind of structure. 05:21.780 --> 05:23.260 Uh, so it's interesting to keep that in mind. 05:23.260 --> 05:28.060 And of course, in week six, we're going to see what anthropic brings to the table in the form of MCP, 05:28.220 --> 05:33.460 which is a different way of thinking about it, a protocol, uh, for, for connecting things rather 05:33.460 --> 05:35.780 than building the actual glue itself. 05:35.820 --> 05:40.940 So I wanted to highlight that and give you that perspective, but it's not going to deter us from getting 05:40.940 --> 05:43.100 deep into Landgraaf right now.