- 1 hour ago
Claude Dashboards gets a first live test in this hands-on review, and the big question is whether it can really compete with Power BI and Tableau. The demo shows how Claude can connect to BigQuery, build a dashboard from survey data, and keep the charts updated as the data changes.
The walkthrough explores how the tool writes queries, creates simple visualizations, and lets you inspect the SQL behind each card and chart. It also tests how well Claude can recreate a Power BI dashboard from an image, then compares the results side by side to see how close the layout and numbers come. Along the way, the review covers sharing, refresh behavior, export options, and the current limits around version control and reproducibility.
If you’re looking for a Claude Dashboards tutorial, an AI dashboard builder review, or a practical look at BigQuery reporting, this video gives a clear first impression of what works and what still needs improvement. It’s especially useful for analysts, data professionals, and small teams exploring AI-powered business intelligence.
SEO: Claude Dashboards review, AI dashboard builder, BigQuery dashboard, Power BI alternative, Tableau alternative, data analytics tutorial, business intelligence tools, SQL dashboard generation, exploratory data analysis, and AI data visualization for analysts and teams.
The walkthrough explores how the tool writes queries, creates simple visualizations, and lets you inspect the SQL behind each card and chart. It also tests how well Claude can recreate a Power BI dashboard from an image, then compares the results side by side to see how close the layout and numbers come. Along the way, the review covers sharing, refresh behavior, export options, and the current limits around version control and reproducibility.
If you’re looking for a Claude Dashboards tutorial, an AI dashboard builder review, or a practical look at BigQuery reporting, this video gives a clear first impression of what works and what still needs improvement. It’s especially useful for analysts, data professionals, and small teams exploring AI-powered business intelligence.
SEO: Claude Dashboards review, AI dashboard builder, BigQuery dashboard, Power BI alternative, Tableau alternative, data analytics tutorial, business intelligence tools, SQL dashboard generation, exploratory data analysis, and AI data visualization for analysts and teams.
Category
📚
LearningTranscript
00:00What's going on everybody? Welcome back to another video. Today we're going to be taking a look at Claude Dashboards.
00:11Now Claude Dashboards was just released on October 8th so it is very new, but it's basically a way to
00:16build dashboards within Claude.
00:18It says right here for Claude Dashboards, you connect a warehouse or CRM and ask a question.
00:24For example, how this week's signups compare with last month's. Claude writes the query, builds the dashboard, and keeps it
00:31current as the data changes.
00:33You can click any number to see its query and each chart shows when its data was last refreshed.
00:39Anthropic pitches it for quick, exploratory questions. Anything deeper goes onto a BI tool.
00:45Dashboards export to Amplitude, Grafana, Hex, Mixpanel, Omni, Perplexity, PostHog, or Sigma with Looker, Monday.com, and Tableau coming soon.
00:56So Claude is pitching this as a tool to get you up and running and creating dashboards very quickly for
01:00exploratory analysis.
01:02Now is it supposed to be a full replacement?
01:05It doesn't necessarily say that. It even says, hey, we can connect you out to these different third-party BI
01:10tools that are meant for more complex dashboards.
01:13But is that something that they may just try to replace altogether in the future?
01:17Probably, if this is any good.
01:19So with that being said, we're going to jump onto my computer. We're going to take a look at Claude
01:22Dashboards.
01:23Now this is not an in-depth review because this is very new.
01:26So I'm just going to test this out, see what it looks like, see what it does.
01:30We'll take a live look at this together to see kind of how it operates and exactly what it does.
01:34And we'll even compare it to a previous project that I did in Power BI.
01:38Now when we come into our Claude Artifacts, you can see we have our dashboards right down here.
01:42It is still in beta.
01:44But all we have to do is click on this.
01:46It's going to open up this window.
01:47It says, what should this dashboard show?
01:49And we're going to have our dashboard right over here.
01:52Now we can connect to outside data.
01:54All we have to do is create a connector for that.
01:56I can say, I have data in BigQuery and I want to connect to it and create a dashboard.
02:07It is then going to prompt me to connect to Google Cloud BigQuery.
02:12I'm going to go ahead and connect this.
02:14So I'm going to show you what this looks like.
02:15It just says Claude Login and then you need to choose how you log in.
02:19I'm going to connect with my Google account.
02:21After I connect to my Google account, I'm going to say continue connecting.
02:23And now right over here, it says we are connected.
02:27Now I just spun this up so that I can upload this data professional survey
02:31because in my other Google Cloud account, I have a lot more data.
02:35I don't want to give it access to everything.
02:36I just want to give it access to this file.
02:39So I'm going to come down here.
02:40I'm going to feed it the path.
02:41I'm going to say I want to bring in this data and create a dashboard with it.
02:49After just a minute or so, it created this.
02:52Now let's take a look at this because this is new to me.
02:55I am looking at this just as you're looking at it.
02:57But it brought in this dashboard.
02:58It said I loaded the survey data of BigQuery and built your dashboard.
03:02All 630 responses are in a new table.
03:04It looks like it created a new table for this in your project.
03:07It removed some of the empty columns because it did need a little bit of data cleaning.
03:11If you go look at the raw file, it grouped some of the data.
03:14And it also created salary.
03:16And it says each range also has a low, high, and midpoint value in dollar signs.
03:21So salaries can be average.
03:23So this dashboard has, and it gives us some information as a headline, the pay, satisfaction,
03:29tool, and priorities, and countries.
03:31And what to know.
03:32Salary averages are estimates.
03:34The survey asked for ranges.
03:35So the averages use the middle of each range.
03:39The charts are live BigQuery queries.
03:41Anyone you share the dashboard with needs the BigQuery connector and access to that table
03:46to see the numbers.
03:47At the bottom, we can ask it to do different things.
03:50And it'll change it for us.
03:51But let's take a look at this survey right here.
03:53It says 66.7% of respondents use Python as their favorite language.
03:57So just a tiny bit of summary.
03:58We have some cards up here.
04:00And as I'm hovering over them, it looks like I can click on them and kind of get the query,
04:05which is very nice.
04:07It says they did a little transformation.
04:09And we also have the results.
04:11So this is really neat.
04:12I'm, again, looking at this live.
04:14But as we click on these, it looks like we can look at each option and how it actually
04:19works.
04:20So I think that's really cool because I know that as we get in here, I would want to change
04:25things and, you know, update and see how things are being calculated behind the scenes because
04:30that's, you know, just what data people do.
04:33So let's come down here.
04:34We have average salary by job title.
04:37Let's just click on this.
04:38Here's the result.
04:39Here's what it's actually looks like it's using to display that visual.
04:44And here's the query that it wrote for that.
04:47So this is really cool.
04:48I think just in general, this is a good overview.
04:52Looks like really simple visualizations.
04:54It's nothing crazy.
04:56But this is, you know, a good looking first pass at a dashboard given that I told it to
05:02just create a dashboard with it.
05:03I didn't give it any other information.
05:05And right here we have data sources.
05:07So we have where we're actually getting the data within Google Cloud BigQuery.
05:11So this is really neat because we can come in here and we can actually see where that
05:16data is, how it's sitting in there.
05:18And we can even update this information and change it as needed.
05:21And so I really like at least my first impressions with the data sources here, how we're actually
05:29able to see the queries behind it.
05:31But now let's ask it to update this and give it a different feel.
05:35Now I'm going to pull this over really quick.
05:37If you guys have taken my Power BI lesson here, this was a dashboard that I created with
05:43this exact data set that we were using.
05:46Now this was more of a teaching dashboard, right?
05:49I'm teaching you how to use different parts of Power BI, how to bring the data in, all these
05:54different things.
05:55So this may not be like a perfect dashboard, but we were able to get a lot of, you know,
05:59useful information out of it.
06:01I want to see if Claude dashboard can change some of the design within Claude.
06:06So I'm going to give it a picture of this.
06:08I'm going to take a picture of this dashboard and I'm going to say, can you recreate this
06:13dashboard?
06:13So I'm going to come right here.
06:14I'm going to paste this in.
06:15I'm going to say, I created this dashboard within Power BI.
06:21I want something just like this with the colors and visuals.
06:29Please create this.
06:30Let's go ahead and give it and just see what it does.
06:33Now it says it's going to rebuild this and you may not have seen it just a second ago,
06:38but it's rebuilding this within BigQuery natively.
06:41And then it's kind of just importing it over here, it looks like.
06:44So from what I'm seeing, it doesn't look like it's building this out within Claude itself.
06:48It's using all of Claude's, you know, AI and intelligence to go into BigQuery and build
06:55it there and then just kind of visualize it here.
06:57I could be wrong on that, but that's at least what I'm seeing while we're doing this kind
07:02of demo.
07:03All right.
07:03So let's take a look at this.
07:05It says, I rebuilt your dashboard to match the Power BI version, same layout, same colors
07:09and same seven visuals.
07:12And just glancing at this, let's take a look.
07:16This looks pretty decent.
07:18Of course, you know, this is a different format or a different layout than the actual one that
07:25we uploaded, which was in Power BI.
07:27So this isn't going to be exactly perfect, but let's take a look at just some of the data
07:33and see if it is similar.
07:35We should have basically the same numbers.
07:37I mean, it is the same data, but we have 4.27, 5.74, 630 count of survey takers.
07:43Let's see if I can scroll over.
07:45I can't scroll over.
07:47Let's see if I get rid of this so I can see this better.
07:49We can.
07:50So I can see it a bit better when I get rid of it.
07:53The numbers right here for difficulty to break into data do look accurate.
07:58There's just some overlap here.
08:00It's, you know, harder to see, I guess, with this compared to the Power BI.
08:04But overall, I think this is pretty similar.
08:08This is about as similar as it's going to get with just one-shotting it by an image.
08:12So I think this is pretty impressive just by itself.
08:15I'm just going to come in here.
08:16I just want to see.
08:17I'm just curious how it actually wrote this out.
08:20And so we're just grouping on this.
08:22It looks like we're pulling it from the different data.
08:26I think it split it out at least when it pulled it in from BigQuery or split it out into
08:30different
08:30tables.
08:31And then it's just rounding it.
08:32And that's perfectly fine.
08:34So this looks good.
08:35And then this is the data set that we are left with to actually create this.
08:39Overall, I am pretty impressed with Claude's ability to replicate this.
08:44It, again, looks like it's being done all within BigQuery.
08:49It does mention where it differs from our Power BI version, which I think is really useful.
08:54And some of the changes that it made.
08:55So you can come in here and you can at least understand and see what it's doing.
08:59I like that it documents all of that.
09:01Let's just say that this is done.
09:04We like how this looks, even though, of course, there's some work to be done.
09:07At least formatting within Claude, it may need some work, right?
09:10But let's just say we want to share this.
09:12Looks like we have some different options.
09:14We can invite people by email.
09:15Now, remember, it said earlier that in order for someone to see this dashboard, they're
09:19going to, one, have to have access or have that artifact connector to that BigQuery, just
09:25like you did.
09:26So if they don't have access to that data within their BigQuery account, they won't be able
09:29to see this.
09:30And so that is a good protective layer.
09:32I like that they have that.
09:33That's pretty standard in almost, you know, any BI tool.
09:37And then next, we have export.
09:38So we can export this to Amplitude, Grafana, Hex, Mixpanel, Monday, and a few.
09:45Other options.
09:46It said things like Tableau are coming soon, or at least I think it mentioned that some
09:51of these other connectors will be coming soon.
09:53That is at least the biggest downside I've seen so far is that, hey, if I want to go put
09:58this in a tool for a client, right?
10:00Or if I want to put this somewhere that I can pass this off to somebody a lot easier, then
10:06there's really limited options for these connections.
10:09Now, I do have some other questions that are just, you know, they're eating away at me.
10:13Like, I need to know these.
10:14One is, let's look at data refresh.
10:17Looks like we have right up here.
10:18It says data as of 16 minutes ago.
10:20Run every live query again.
10:22So I can refresh this.
10:23And it's going to query off of the actual data.
10:26And it looks like it is refreshing.
10:28So it looks like we have a live connection to our data in BigQuery.
10:32That's always good, right?
10:33I need to know.
10:34And I need to be able to see when this data was updated and things like that.
10:38So I think this is fine, although I might want to see a history of when this data was
10:42refreshed and things like version control.
10:44Is there version control in here?
10:46So I'm just going to say, is there version control in this dashboard?
10:52I'm just going to ask it because genuinely, I'm not 100% how to, I'm not 100% sure how
10:57to
10:57find this.
10:58It says there's partially.
10:59Here's what I know from how this dashboard type works.
11:02It says the queries have history.
11:03So in the sources panel, you can edit the SQL behind the chart, test it, save it, and
11:08restore an earlier saved version.
11:09That's the main version control the dashboard gives you.
11:12The layout and visuals don't have a restore button that I know of.
11:15And that's something I was curious about because if I go back up here to a previous
11:19version, this was the first one it created.
11:21If I click on this, it's just the same one.
11:24This doesn't actually revert to the previous one, the first version that it created.
11:28It's kind of overwriting itself.
11:30And I don't see that version control where I can go back and take a look at a previous
11:35version.
11:36I would probably have to export that to a third party tool like we were looking at up
11:41here.
11:42I'd have to export it to one of these.
11:44It does say though, your first design isn't lost.
11:46I still have the code for both versions from this chart, the first card based layout and
11:51the current Power BI style.
11:52For lasting history, I can save each version as a file.
11:56So you can do that.
11:58In fact, I can't imagine that they wouldn't be able to easily put a simplified version
12:04of version control in here for this.
12:06I'm sure someone's already thinking about this and building this out as we speak.
12:10So I'm going to say, yes, I want you to create a type of version control for this.
12:17So it doesn't have a built in version control and that's okay.
12:21We're just trying this for the first time testing it out.
12:24I didn't have like crazy high expectations that I would have every feature that would
12:29be within a standard BI tool.
12:31But what I will say, and while it's doing this, is I will say overall, this is going
12:37to be really good for people just getting into their data.
12:41Now, I have a lot of questions and some concerns as it comes to reproducibility.
12:46It's running these things within BigQuery.
12:49It's saving the tables.
12:51It's all, it's basically doing this within BigQuery.
12:53If I took these exact queries, like if I came in here and I took this query, I need to
12:59dive
12:59into this more.
12:59If I took this query and I ran it, and especially as we get more advanced, would I be able
13:04to
13:04replicate all of this within BigQuery?
13:07Would I be able to replicate this in other BI tools?
13:10It's very possible I could, but I need, as an analyst and as a data professional, you
13:16need to be able to validate your data, have it be reproducible, and it be consistent across
13:22the board with whatever rules you set in place.
13:25And so these are things that I would be slightly concerned about, but I think it's, in general,
13:30it's one-shotting this pretty decently.
13:32And so back to the version control, it looks like it has uploaded the zip for me.
13:38I'm not going to actually download this, but this is the read me.
13:41We have a version one and our version two, and we can put this all into Git.
13:45So we can, we can pseudo create our own version control here with using Git and probably GitHub.
13:53But if this is all local, just use Git, use it on your computer.
13:56It'd be super easy to do.
13:58And so if you are a working professional and you know how to use Git and you know how to
14:03use, you know, GitHub, if you want to use that, then you should be able to do this,
14:07but it doesn't have it built in and that's fine.
14:09I'm, you know, we're just messing around here.
14:11So that is our first look at claw dashboards.
14:15In general, I think it did really well.
14:17I mean, there's a lot of tools out there right now that can do something very similar.
14:22And I've reviewed some of those on my channel.
14:25I think that, you know, the purpose of this is to get quick insights into your data.
14:30I don't know if this is going to be like a go-to place for teams, right?
14:36Actual data teams to go and build things out.
14:39There are going to be some limitations within something like this, right?
14:43If you actually want to share this with clients and you want to share this with people, you
14:48have to get it into a third party tool anyways.
14:50This is a good place to maybe build it out initially and then put it in there.
14:54But a lot of these tools are getting this type of functionality anyways, where you can say,
14:58hey, AI, here's what I want.
15:00Here's kind of what I want it to look like.
15:02Build this out for me.
15:03And it gives you a good first version.
15:04And within those tools anyways, you're going to be able to validate those queries and those
15:09outputs.
15:10In general, though, I think this is really nice for a lot of people in a lot of small
15:14businesses.
15:15I don't think this is going to be used at like big companies.
15:19There's just a lot more that's fit in to a full-blown BI tool that is just not in here.
15:24So as a first look, as you are looking at it live, as I was just exploring
15:28it as well, I think this is really good.
15:30You can upload CSV files and you can upload, you know, different connections to different
15:35data sources, even in Databricks and Snowflake.
15:37And this is going to be really good for a lot of people who just want to take a quick
15:40look at their data.
15:41Now, as they come out, as they improve this with things like version control and maybe
15:44different ways to share it just in general, right?
15:47Maybe I'll be able to download this and not have to have some private URL and you have
15:51to have access to the data.
15:52You can just look at a static dashboard instead of it having to be connected to the data source.
15:57So there's so many different things, even just off the top of my head that I can think
16:01of that it doesn't have, but I am impressed.
16:03I do think this is useful.
16:05I do think that this is good.
16:06And I think I should, you know, mess around with this more and give a kind of a tutorial
16:09on how to use this once I dive into it more and I know a lot more about it.
16:14So in the end, my review on it is that it is pretty good.
16:18I give it a solid eight out of 10.
16:19I just don't think it's a full blown BI tool.
16:22It is not a replacement for that at all.
16:24If you've been using Power BI for a long time, you know that there are so many things that
16:30kind of are behind the scenes that are not just connecting to the data and building the
16:34report, right?
16:35That is important, but there's so many other things that go into it.
16:38And so eight out of 10, I think is really solid for just being in beta.
16:42As they make updates to this, as they make improvements to this, as I get my hands on it,
16:47I'll give a more in-depth review and probably some type of walkthrough on how to use this.
16:51With that being said, I hope that you enjoyed and learned something from this video.
16:54If you did, be sure to like and subscribe and I'll see you in the next one.
16:57Bye.
17:00Bye.
17:09Bye.