Claude Just Built a FULL Data Dashboard (No Coding Needed) 🤯

Key Takeaways

Hello everyone and welcome to this video. In this video, I wanted to to use Claude chat to perform data analysis and data visualizations and also create interactive dashboards. So let's go ahead and get started. So what I've got here is I got a prompt and basically I'm asking Claude chat to create 10 different data visualizations to highlight various aspects of this data. I want it to include interactive charts and here the plan is to upload what we call it the cancer data sets. So this is a standard data sets that we simply use to train machine learning models. Let me show you. Please note that I included this data as well as part of our course package. So simply here I got data related to breast cancer and here we got a lot of features. So we got for example the mean radius, mean texture, the mean perimeter and if you scroll to the right here, you should see we have the target column and this is simply a binary column that either shows zero or one meaning that either every row has is a patient and based on their features, we can predict whether they have cancer or not or whether the tumor is malignant or benign. So really all I need to do is I'm going to use that prompt and I'm going to upload this data to Claude here. So I'm just going to paste the prompt and you can see here on the bottom left corner, there is a plus icon. So now you can add files. You can also connect Claude as well to let's say your Gmail and have a connectors and so on. So I'm just going to click plus and I'm going to say add files or photos and I'm going to navigate here to my data. So I'm going to head back to So this is the data sets here and here I got that cancer data set. It's simply an Excel sheet that includes here this cancer data. And please note that I'm going to enable Opus 4.6 with extended thinking as well. So let's go ahead and run it. So now you should see basically Claude is going to first read the data. It's going to because I asked it to create an interactive basically dashboard. It's going to write code and it's going to show you as well or render that code on the right hand side. I actually ran it before and it was incredibly powerful. You can also see it was able to read the front end design skill. So actually going to create a hopefully, okay, if it works similar to what I ran it before, it's going to show you like an amazing dashboard because it has already used one of these front end design skill. Please note that with skills, think of it as you take the Claude and kind of make it a super LLM or super large language model because now you can have the ability to create front end design. You can also use it to interact with PowerPoint presentations by getting that PowerPoint skill. You can also ask it to create Microsoft Word documents. Again, these LLMs you can extend them with skills and plugins to take that to the next level. So now you should see it's actually creating the comprehensive interactive chart and look at this, actually writing that code behind the scenes. It's able to extract all the different numbers for radius, area and you can see the target here as well, either malignant or benign. If you remember that zero or one dimension, it's actually like showing it to you here and showing you as well kind of the area radius and if you go up, you should see what it's doing. These are all the different for example colors that has been set and these are the different visuals that are going to be used, right? So you can see for example it's going to do a radar chart, it's going to do a scatter chart, it's going to do line chart and you can see that these are all the different charts here that are being created as part of our dashboard. So what I'm going to do is I'm going to let it run and I will be back once the dashboard is created. All right >> [clears throat] >> and here we go. Now you should see, I would say like a next level dashboard here that has been created showing you how many samples in total. So you can see almost 569 samples. And please note that this is a benchmark data, so I'm very familiar with the data and these numbers are quite right. I can actually show you here that 62% of the data is benign, 212 is malignant. There are 30 features. You can also see as well this interactive chart. So you can see the class distribution here. It's showing you as well kind of the distribution of both the malignant and benign cases. You can see as well if you scroll down, see like look at these incredible charts. I think I'm like amazed by the power of Claude, of course with the ability of writing or using these front end skills as well. You can see here we got a nice box plot. You can see as well here a nice scatter chart as well. And please note that this simply is an interactive chart and of course if you want to see what is happening behind the scenes, you can see the code that has been generated here as well. One key point that I would like to share with you is you will find that simply Claude when it's running, so it took maybe like a few minutes for it to run. I believe it might have the time here. So I would say at least five minutes it took and basically what you notice is that these LLMs right now, they have that sort of self reflection loop. That means before it generates any output, it actually goes back and evaluates its own work and that's why you will find that these dashboards are actually quite quite accurate. You know, if you if you if you go here, you will find it looks looks pretty good. Looks like a pro designer actually created it. You will find for example that it actually here found that there was an issue and it's telling you let me fix chart 10 and there's an issue here with it. So it went back, it fixed it, it came back here, it did another some sanity check and now it came back with a dashboard with 10 interactive visualizations telling you done. And if you're wondering, okay, like are these models actually like smart or not? This is a website that can actually tells you the capabilities of these LLMs or large language models. So just quickly I wanted to share that older models, especially if you have used maybe AI back in 2024, these let's say Opus model, the older Claude, it used to finish tasks that used to take humans two minutes only and now they break. So for example, they can maybe fix bugs in a small Python library or maybe do something very basic. You can't let them go and work for example for an extended amount of time. But then what happened around the end of 2025, something crazy that like basically everyone who works in the AI field is kind of amazed or astonished by is now these AI models. Now let's say GPT-5 and then followed by Opus 4.5. Now they can do or work for let's say four hours. Now they can finish a task that used to take humans four hours and now they can do it autonomously. And you can see that here when we created our dashboard, actually went there, well it found a mistake, it came back, it fixed it, it went there, it came back and you can let it run and do massive tasks and then suddenly Claude Opus 4.6 which is the model that I'm running right now, now it can literally work for 12 hours. So it can finish a task that takes humans 12 hours and now it can do it autonomously. Please note just to be clear is that this is here with a 50% success rate. So like 50% chance it's going to finish it. If you want to increase the chances of success, let's say 80%, now you will find Claude Opus 4.6, now it can actually finish a task that takes humans let's say one hour to complete with 80% success rate. But just with that rate, with that exponential kind of insane growth and and and improvement in capabilities, we are expecting over the next year or so, you will find the next Claude Opus 5 or maybe 4.8 or something, it's going to be like you can just give it any task and it's going to finish it with almost 90% success rate. And with that, that's all I have for this video. I hope you enjoyed it and see you in the next one.