Episode Categories:
Resources:
Join the discussion: Got something to say? Let us know on LinkedIn and network with other AI leaders
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Start Here Series in our Inner Circle Community: Join for free access
Unlocking AI Productivity: Anthropic’s Claude Interactive Apps for Business Leaders
Enterprise decision-makers often find themselves patching together multiple AI tools to achieve workflow automation, data enrichment, and cross-app collaboration. This patchwork, dubbed “Frankenstacking,” creates hidden inefficiencies and bottlenecks. The recent release of interactive apps within Anthropic’s Claude platform is shifting this landscape by offering a practical, unified approach to handling complex workflows—all from inside a single AI interface.
Claude Interactive Apps: Workflow Integration Without Duct Tape
Unlike most AI integrations that simply deliver text-based transcripts or static outputs, Claude’s new suite of interactive apps introduces real-time, multi-app workflows capable of sharing context and operational control. Business leaders using SaaS tools for project management, analytics, design, communication, and customer support can now orchestrate multiple services without manual intervention or incessant copy-pasting.
For example, when tasked with vetting hundreds of podcast guest submissions, a workflow within Claude pulls the last five pitches from Gmail, enriches each with company data via Clay, and custom-builds a branded presentation using Canva—all within a single prompt. This cutting down of manual research and context switching translates directly into time savings and sharper decision-making.
Practical Connectors: Pinpointed Business Value in App Selection
Getting started is straightforward: select and activate the Claude connectors (the apps Anthropic calls interactive tools), currently numbering about ten, including Asana for project management, Clay for data enrichment, Canva and Gamma for design, Slack for communication, and Box for document management. Each connector offers specific capabilities, such as “find and enrich company” or “add contact datapoints,” allowing businesses to customize their AI-assisted workflows to precise operational requirements.
Importantly, not all connectors are interactive or fully featured; the user’s subscription tiers (Pro, Max, Team, Enterprise) dictate the range of available functions, underlining the necessity for clarity when choosing integrations. Browsing connectors via Claude’s dedicated directory empowers leaders to identify and leverage the tools that best match their existing SaaS profile.
Real-Time Collaboration: The Embedded GUI Advantage
A major advancement Claude brings with interactive apps is the embedded graphical user interface (GUI). Instead of handling information on a text-only basis, the apps present live panels so users can view, edit, and interact with data—such as analytics dashboards from Amplitude, editable slides in Canva, or instantaneous Slack messages. This live collaboration eradicates human scaffolding: there’s no longer a need for repeated manual actions to synchronize information between platforms.
Claude’s model context protocol (MCP) ensures seamless context passing between these apps, enabling rich interactions—like researching companies in Clay and funneling the insights into a presentation generator in Gamma—all managed from a single conversational thread.
Comparing Claude and ChatGPT Apps: Enterprise Reliability and Output
While ChatGPT apps have led in terms of quantity and initial release, Claude’s interactive connectors distinguish themselves by targeting enterprise-grade productivity. Business-centered partners like monday.com, Asana, Box, Clay, Gamma, and Slack are integrated to support robust, cross-functional tasks. The output from Claude’s interactive mode is deeper when it works as intended, but users should anticipate early-stage bugs and occasional mismatches in app selection, as documented in recent workflows switching from Canva to Gamma unexpectedly.
Despite these quirks, the ability to share context and drive multi-app actions signals a mature step in AI operating systems, rivaling traditional connectors and apps found in ChatGPT and Google Gemini.
Addressing Security and Control: Data Permissions Matter
Leaders should note that enabling connectors requires explicit data permissions. Authorizing read and write access—such as sending messages from email analysis directly into Slack—demands careful attention to company security policies. Businesses must ensure compliance before leveraging these capabilities, particularly when integrating with sensitive communication or data storage platforms.
Seven Pinpointed Use Cases for Claude Interactive Apps
From the episode’s segment and accompanying guide, Claude’s interactive apps deliver strong value in:
Vetting guest pitches with CRM enrichment and presentation automation
Summarizing and broadcasting cross-department feedback directly into team channels
Instantly building project timelines from chat analysis into Asana
Surfacing document insights from Box into open conversations
Creating branded slide decks with design apps using accumulated context
Assigning task updates across SaaS workflows without manual handoffs
Visualizing analytics and performance dashboards embedded directly in the AI interface
Each of these workflows shrinks human effort, increases throughput, and enables more strategic allocation of attention to high-impact areas.
Claude Interactive Apps: Immediate Impact, Long-Term Promise
While the interactive app system is not yet flawless nor exhaustive in its integrations, its practical business impact is measurable: reducing context switching, automating cross-app tasks, and operationalizing AI for end-to-end workflows. Organizations relying heavily on project management, design, communication, or analytics SaaS tools will find immediate value in consolidating their day-to-day through Claude’s evolving platform.
The actionable insight: businesses positioned to benefit should begin testing and implementing these workflows now, as current usage delivers a competitive edge and prepares organizations for the next wave of advanced, agentic AI operating layers.
For detailed guides and workflow samples based on real-world use cases, refer to resources like the Everyday AI newsletter or connect with active practitioners leveraging Claude’s interactive apps. Early adoption remains the clearest path to scalable, context-rich AI productivity.
Topics Covered in This Episode:
- Claude Apps Interactive Tools Overview
- Anthropic’s Model Context Protocol (MCP) Explained
- How to Enable Claude Interactive Apps
- Claude Apps vs. ChatGPT Apps Comparison
- Workflow Automation with Claude Connectors
- Key Claude Apps for Productivity (Clay, Canva, Gamma, Slack)
- Real-Time Collaboration and Embedded Interfaces
- Limitations and Bugs in Claude Interactive Apps
Episode Transcript
Jordan Wilson [00:00:18]:
Most AI tools that analyze sales or customer support calls turn those conversations into a text based transcript, but text only transcripts miss most of the value, and Modulate fixes that. Modulate's new Velma voice native AI model and their ELM technology actually understand what's happening on those calls. It picks up the valuable tone, timing, emotion, and intent that all AI transcription tools can't provide. So whether it's for sales, customer support, or voice agents, Modulate's new Velma model helps you capitalize on what text only AI tools miss. Demand more from your AI today with Modulate at modulate.ai. I'd say for the last few years, business leaders who are AI native and using large language models regularly, kind of doing this thing I'll just call Frankenstacking. Right? They're piecing together these different parts of processes that different AI tools are pulling together. So it's kind of like this duct taping of a workflow that we didn't have to do before chat g p t, but now we do.
Jordan Wilson [00:01:32]:
But that's changing. And if you've been listening to the show at all, you're probably tired of me talking about AI operating systems, but it's something that I've been personally very, bullish on for years. And the most recent release now from Claude just really hammers that point home. And that's what we're gonna be going over today. Claude's new interactive apps and their interactive tools and how this can really increase your productivity and maybe replace a lot of that context switching or frankenstacking that I think humans that have been using AI correctly that we've been doing for the past few months. Because now, well, these apps inside of large language models like Claude can hopefully do a lot of that for us. Alright. I'm excited to dive in.
Jordan Wilson [00:02:23]:
I hope you are too. Welcome to Everyday AI. If If you're new here, this thing, well, it's for you. It's for everyday business leaders who are too busy to spend hours every single day trying to figure this AI thing out. So that's what our daily livestream podcast and free daily newsletter do for you. We help you grow your company and your career. So if that's what you're trying to do, it starts here with the unedited, unscripted livestream podcast. But to take it to the next level, you gotta go to our website, youreverydayai.com.
Jordan Wilson [00:02:51]:
There, we're gonna be recapping the highlights from today's show as well as the AI news that you need to know to stay ahead. So this is our kind of weekly segment. We've been doing this for almost a year now, called AI at work on Wednesday. So on Wednesdays, we kind of go over, new tools and features from the big four. That's Google, Anthropic, Microsoft, and OpenAI. So today, we're gonna be looking at Claude, and they've honestly been on kind of a terror lately. Right? So, obviously, toward the 2025, Claude Code just kind of exploded in popularity. And then we got the kind of nontechnical version of that, which we went over last week in Claude Cowork.
Jordan Wilson [00:03:33]:
We've had all these kind of spin off, projects. You know, there's the Claude bot, which is now called the Molt bot. Right. But this one, I think, could be as big as all of them. So it hasn't even been out for a full forty eight hours yet, but now, Anthropic has released these new interactive tools inside of Claude. So we're gonna do it a little different today. We're gonna start live. Alright.
Jordan Wilson [00:03:57]:
So if you ever wanna see the video version of this, if you're listening on the podcast, you can find that for free on our website, youreverydayai.com. So livestream audience, if you could let me know. If you could see my screen, that would be super helpful. See if I can make myself a little smaller there. There we go. So I am just inside of clock. Alright? And I'm gonna go over exactly what we're doing, but, hopefully, this is gonna work live. I have some other examples that are, kind of done because they take a little longer.
Jordan Wilson [00:04:24]:
And full disclosure, this has been buggy. This has been a little buggy since it started, but when it works, it's actually really good. Alright. So I'm gonna go ahead and show you. So when you go to claud.ai, and log in to your account, you have to make sure that you, enable any of these kind of new connectors. So it's weird. There's kind of this, like, naming mismatch going on because they're technically kind of apps, and that's what most people are calling them. But they're under connectors, in your, kind of toolbar, but, Infropic is also calling them interactive tools.
Jordan Wilson [00:04:57]:
Alright. So all you have to do is click the plus icon when you go into your cloud account. Click connectors, and then go to manage connectors. And then you're going to go to browse connectors. So there's all of these. There's no way to filter them to know which ones are the interactive ones. But if you click, right so I'm gonna go ahead, one that I know is interactive and I already have installed, Clay as an example. You gotta scroll down, and then there's a capability section at the bottom, and then you can see it's interactive.
Jordan Wilson [00:05:26]:
So Anthropic did list, I think, like, eight or nine on their website, but I went through all the connectors manually, and I noticed that there's some that are interactive that they didn't even mention. So, it's a little confusing to even get these set up in the first place. But if I'm being honest, it's kind of confusing in Gemini, in chat GPT as well. Right? Chat GPT even renamed them. So that's how you do it in Claude. Click the plus button. Go to connectors. You're gonna have to find the individual ones.
Jordan Wilson [00:05:55]:
I think there's only about 10 of them right now, and I'll tell you which ones they are. And then you're gonna get going. Alright. So now I'm gonna go back. Hopefully, my, my my prompt is still gonna be there. Perfect. So, no. No.
Jordan Wilson [00:06:09]:
That's not the right one. Fantastic. So, I'm gonna have to grab the correct prompt. Let's do this one and paste it in there. There we go. So you need to make sure that you have the correct connectors active when you are testing this out and when you see how this works. So I'm gonna get this started, and then I'm gonna explain, what's happening here. And this is a real use case.
Jordan Wilson [00:06:33]:
Right? I think a lot of people always ask me, Jordan, how do you use all these AI tools? And I wanna know how to use them. Well, this is how I use them. So here's an example. I said, carefully look at my Gmail. So it's using my Gmail connector. So you can, use these new apps, but also you can pass information from a non interactive connector to an interactive app, if that makes sense. So the Gmail is not an interactive app, but it's a connector. So I'm telling it, look through my Gmail and five find the last five guests who have pitched to be a guest on the Everyday AI podcast.
Jordan Wilson [00:07:09]:
Then use Clay, to enrich with basic company information. So Clay is a a data enrichment app. So I'm saying then use Clay to enrich with basic company information so I know more about the companies. Then present all of that data visually in a Canva presentation in a minimalist Apple esque design focus. Focus specifically on what you know about everyday AI. Right? It has memory of me, Claude does. And the value each guest could bring to our audience and give an audience fit score out of 10. So, we get more than a thousand guests each year to pitch to be on the show.
Jordan Wilson [00:07:46]:
One thing that I struggle with is I don't have enough time, to manually research them all. So I'm constantly looking at workflows like this that can do a lot of it for me. So Claude obviously has a lot of my information, what I look for, in in the memory. And so here, I'm having it go look deeply through my email. I'm giving it some personal context, to start with. And then, saying go find these last five people that have pitched. Go find their companies. Use Clay to know more about these companies because half the time I don't know because there's, more new AI companies every day than there are, you you know, I don't know.
Jordan Wilson [00:08:22]:
Fish in the sea. There's thousands of new AI companies every day. I swear. So half of the people that pitch me, I don't know, but they could be legit. Who knows? Usually, I just spend a lot of time manually doing this, and then I have to read through a long sea of text, which I hate. So this is a use case where it's, kind of that frankenstacking that I've been doing a lot manually. Right? So, unfortunately, Claude right now, doesn't have a way to schedule tasks or schedule prompts like you do in chat GBT. I hope that they add it.
Jordan Wilson [00:08:51]:
Gemini has been rolling this out slowly as well. So this is an actual use case. So I want you to think first, what are those roles that you are duct taping right now? That you're franken stacking. Right? Yes. Even if you're using AI to its fullest, a lot of times you're still doing all of this context switching, copying and pasting, using, you know, going from these different apps that you use, these different SaaS applications that your company relies on, and you're constantly having to just patchwork everything together. So that's the great thing about these new apps inside of large language models is the ability to pass the context. Alright. So we're gonna go back and check on this a little bit later.
Jordan Wilson [00:09:29]:
Alright. But let's get back to kind of the the the details. Right? We just went straight into it. So, they did launch with nine apps, and they are interactive, covering everything from analytics to project management and design. So the open standard is really what pull like, pulls this all together. Right? So this uses Anthropic's model context protocol, which it is really impressive. Right? And you'll see here in a minute, and I'll try to describe it for our audio only audience. But, essentially, this allows you to both, share data from all of these different sources without you having to duct tape it altogether, but also to, in certain instances, interact with a graphical user interface or a GUI.
Jordan Wilson [00:10:15]:
Right? And so we're also gonna be covering the different apps, how the MCP protocol kind of powers all of this, and then also compare them to chat GPT apps as well. So here's the overview. You have to be on a paid plan right now to access this. Alright? So that includes the pro, the max, the team, and the enterprise. And at least right now, it's only available on the web and desktop. So, they did say that these, new apps will be rolling out to Claude CoWork at some point. I'm excited for that. I've been, really enjoying using Claude CoWork, which is kind of the nontechnical or graphical version, of Claude Co.
Jordan Wilson [00:10:56]:
So here's how you enable them. So aside from the way that I went to, there is also a URL. We'll put that in our, in our newsletter. It's a little faster to get to in all honesty. You just go to claud.ai/directory, and then like I said, you have to go through and find the ones that are interactive. And then the features are limited. Right? Depending on what plan you have. So as an example, I'm on a free, I'm on a free Clay plan.
Jordan Wilson [00:11:21]:
I don't subscribe there, but I have an account. So I can only take advantage of, once I authorize Claude, to essentially read and in some cases, write, across the two different apps, I can only use features that are available. Oh, one one thing that I should, I I I should denote here, and, I'll I'll I'll just kind of go back over, into live mode. Here we go in a second. So for all of the different apps, you might not even know what they do. Alright? So, there is a nice way to do that. So I'm actually gonna go to claud, dot a I slash what is it? Directory. There we go.
Jordan Wilson [00:12:02]:
There we go. So, as an example, I'm gonna go to Clay. So if you scroll down, you will actually see the different tools that you can use, in the interactive mode. So, for the most part, you should be able to understand the capabilities, that these kind of new interactive apps have. Sometimes you can just ask them. Sometimes it doesn't work. One thing I I I do not like this about anthropic claw. You can never, like, at mention something.
Jordan Wilson [00:12:35]:
Right? So you just have to have all your, kind of interactive apps, enable, and then you just say, hey. Use Canva or, you know, use Clay and go do this. But I found oftentimes that doesn't even work. I actually really prefer, the version like in chat g p t where you at mention. So you would type, you know, at clay or at canva, and it pulls it up like you would in a team's, Slack message, etcetera. So you can see, oh, okay. I am specifically sending this command, and the system knows I'm sending this command, to the actual app. Right? So Infropic always does it a little differently, so I'm not a huge fan of it.
Jordan Wilson [00:13:10]:
It's the same thing with their skills. Right? You know, Infropic says that Claude decides when it uses the skill. So you can tell it to, and sometimes it will even when you don't want it to. I like having the control, but I get nontechnical people might not like this. But, as an example here, I can see the different tools, or the different capabilities that Claude or that this, Clay app has. So as an example, it can, find and enrich contacts at company. These are just tool called names. Right? But for the most part, they're descriptive, and I can understand.
Jordan Wilson [00:13:40]:
Find and enrich list of contacts, find and enrich company, add contact data points, etcetera. So these are all things that if I'm like, okay. Is this company that pitched me, are they legit? Are they not? Normally, I would, myself or an agent, have to go do, you know, some of this, research work. So these are the examples you can always see, what the different apps are capable of if you go and click and check that out. Alright. So getting back to, kind of our details on how these new interactive apps works. Like I said, the features of the apps might be limited by the plan that you have. So here's why it, I think, matters and kind of what they can do.
Jordan Wilson [00:14:20]:
I kinda broke this down into three different, main functions that apps can do. And, again, interactive apps, interactive tools, whatever you wanna call it, but these are different than connectors. Right? The biggest thing is embed. So it's embed, act, and sync. So embed, the apps, depending on which one you use, some of their graphical interface is going to actually show up as interactive panels. Right? Same thing in chat g p t, but you're gonna see usually a slimmed down simplified version of if you were using, that software or that app on their actual website. Right? It's usually not as feature rich. I actually think the chat g p t version is is better in that regard, much better.
Jordan Wilson [00:15:04]:
But you can actually see and interact with those elements. Next, you can act. So you can view, edit, and even take actions without leaving the conversation, and I'll show you some examples of that. And then sync, like I said, two way. Get rid of the human scaffolding. Right? Which is what a lot of us have been doing. Right? Like, AI actually because it saved us so much time, but the AIs didn't always talk to each other. It actually created all these new kind of AI assisted human tasks that didn't exist pre chat gbt.
Jordan Wilson [00:15:33]:
And this, kind of the two way sync, I think, does get rid of a lot of that. So here's some of the, the apps that were announced, and some of the capabilities. So, let's just go let's let's go straight, straight into it here. Here we go. So we have Amplitude, which helps you build analytics, charts, and Claude, Asana project management. So you can turn chat chat conversations into projects, tasks, and timelines. Box, the storage company, actually been super impressed with Box. They're always, you know, partnering with all the different AI companies, really at the forefront, I'd say.
Jordan Wilson [00:16:08]:
So you can, in your Box storage, you can search files, preview documents in line within Claude, extract insights, and ask questions about the content. Canva, so you can design presentations in Canva and see. Although that part has been iffy for me, it seems like it functions different. I did the same prompt, in the Canva interactive app, like, seven times, and I think I got three very different, displays. So very strange at times. Like I said, I don't know if it's early bugs or if I don't know. You just need to prompt a little bit different. We'll find out over time.
Jordan Wilson [00:16:41]:
Clay, like I said, research companies find contacts. Salesforce and Agent Force three sixty are coming soon. There's also hacks, monday.com, Slack, Figma, for design. Another one I found that wasn't listed on, Anthropic's kind of launch page, which is a good one is is Gamma. So didn't even know Gamma was an interactive app until I, you know, spent a couple hours poking around and looking at them all manually, and I'm like, oh, this one's interactive. Cool. Alright. So that is, kind of how this works.
Jordan Wilson [00:17:16]:
So, a couple big differentiators and what this means. So I think a lot of people look at large language models as giving you text answers. Right? But now, like I said, the ability to have that embed act in sync, you know, Claude isn't just giving you answers anymore. It's giving you a live visual interface that you can see, edit, and interact with, and it does become this operating layer. Right? This is, again, truly bringing those day to day, manual, work processes that you're doing in all of these other apps and softwares into Claude and sharing the context. That's the big thing. Right? Being able to pass in my example, we'll see if it works. Right? But being able to pass the context of Gmail without me having to do it straight into Clay, you know, and then Claude will power Clay to go find all this information, and then Clay will pass that straight off to Gamma as in my example.
Jordan Wilson [00:18:07]:
Right? So it does become this operating layer, that's operating all of these different services. So it is a shift. So previously, Claude had connectors. And for the most part, there was no graphical interface, number one. And the it was kind of limited in terms of what you could do on the right side. Right? Like, actually changing files, taking actions. Obviously, with, Anthropic's custom MCP servers, you can do a lot of that, if you wanna get, you know, a little bit more technical. But now it's just in real time.
Jordan Wilson [00:18:40]:
You can collaborate with Claude in real time and see the work that it does on the screen. Alright. So let's compare Claude apps versus chat GPT apps in just a second after a quick word from our partners. Fraudsters used to need fake documents or stolen credentials to scam a business. Now they just need a few clicks to get your CEO to say anything they want. Voice deepfakes rose more than 680% last year. And most fraud detection systems only flag suspicious transactions after the money is already gone. They're not actually listening to the call where the scam is happening.
Jordan Wilson [00:19:18]:
Modulate is. Modulate's new Velma model analyzes live conversations for the signals that give fraudsters away. Stress patterns that don't match the story they're telling, urgency that sounds performed instead of real, voices that are synthetic instead of human, all detected for your business in real time, not in a report days later. Modulate's Velma model was trained on twenty one billion minutes of real audio and is trusted by Fortune 500 companies. It outperforms voice models from leading AI labs, and it's a 100 times more cost effective. So go see Velma catch what your current tools miss at modulate.ai. Alright. So, let's get back to it real quick actually before we, before we go into the comparison.
Jordan Wilson [00:20:08]:
So, the original prompt I started with, which is why I can't do all of these live, is actually still going. So, we'll we'll get into it a little bit more, but, I think it's been at least ten ten to twelve minutes. Alright. So let's compare the new and this for easy naming, let's just call them the, the Claude apps versus chat g p t apps. So chat g p t actually beat them to the punch, with with this one. Right? Which is, pretty pretty impressive, I'd say. And the reason why I think it's kind of impressive that chat g p t actually came out with the apps first as well because OpenAI's apps SDK, so on the developer side, that allows developers to build these type of apps is actually built, on Anthropic's model context protocol. So it was impressive to me that OpenAI kinda beat them to the punch.
Jordan Wilson [00:21:04]:
So OpenAI's, chat g p t apps launched, in October. At the time, we're only seven apps, and I didn't think that they were that great. But since then, they've actually really doubled down on them. I think there's more than, 40 apps right now. So, pretty impressive, a much wider variety right now, than Claude. So I will say this. Claude's partners maybe skew a little bit more enterprise. Although, obviously, ChatGPT apps have a lot more of that as well.
Jordan Wilson [00:21:36]:
I don't know. Maybe it's because, anthropics, these new interactive tools, interactive connectors, interactive apps, whatever you wanna call them, are a little newer. I'm sometimes more impressed with the output, but it's way buggier, and it's way less consistency. So, I'd say overall, it's kinda like I'd say like the the the Apple Android experience, right, from ten, fifteen years ago. You know, everyone's like, oh, if you just want something to work, you know, use an Apple. I I I kinda get that when using ChatGPT apps, versus using cloud apps. When they work, they're fantastic. But like I said, they're very buggy, and I've run into very few bugs on ChatGPT apps.
Jordan Wilson [00:22:18]:
So kind of pros and cons or maybe it depends on what your business, what your company uses in your day to day workflow, that might, you you know, help you understand. And something to know as well. So Chad GPT actually kinda rebranded their connectors. So they separately, like Claude still has this separate thing. They have their, you know, non interactive connectors and then their interactive connectors. So, essentially, in December, which a lot of people missed, OpenAI just kinda got rid, quote, unquote, of connectors that they rebranded them now as apps. So, some of Chargebee apps don't actually have that interactive element. Some do.
Jordan Wilson [00:22:57]:
Right. So that's the difference, some of the similarities and differences. Alright. So that's all for our boring bullet points. Now we can get back in, and let's do this live. And let's see. Alright. Cool.
Jordan Wilson [00:23:11]:
We finished. So let me exit out of this here first. So let's go into just that one that we did. So as you remember sorry. As a reminder, here was the original prompt. Carefully look at my Gmail and find the last five guests who have pitched to be a guest on the Everyday AI podcast, then use Clay to enrich with basic company information so I know more about the companies, then present all of that data visually in a Canva presentation with a, in a minimalist Applesque design style, blah blah blah. Give an audience fit score, all that good stuff. Alright.
Jordan Wilson [00:23:40]:
So y'all, if you know anything about how I do these demos, what's the one thing I say do first? Always look at the chain of thought. Right? Always, you know, go back and check-in AI's work. Never take it, you know, never just take it as truth. So always go through and look at the thought process. I'm not gonna bore you, but, it starts by thinking, then it goes and it searches, through my mail to pull out, some of those guest pitches. And then you'll see right away, there's a new little icon. So if you kind of scroll through what Claude thought, you can go and look. You can always see the icon of the app or the connector that it used.
Jordan Wilson [00:24:18]:
So here we are, Clay. And I can see it used the tool find and enrich company. So maybe I wanted to find and enrich contact. I would have to tell that a little more specifically in the prompt, but I didn't even, tell it which tool to use. I just left it open ended, and, obviously, it did a good job. So here you'll see, it's actually a very nice interface. Right? So, this company, TraceFuse, they must have pitched me recently. I didn't see it yet.
Jordan Wilson [00:24:50]:
So I can see it. It's it's a very nice interactive, display that's brought in from Clay. So, you know, it's almost like I I can scroll over. So it's like a an an embedded website within the Claude interface. Right? So I can scroll kind of up and down this, Clay interface without it moving the rest, of the Claude window, if that makes sense. And there's some tabs up here. I can see the people. I can see the company.
Jordan Wilson [00:25:18]:
Right. So this is all public information. Right? So I don't care that I'm showing any of this. Right? There's a nice, kind of graph here that shows the company's, headcount growth. Right? So I can scroll through, see, okay. This is a big company, small company, all that good stuff. There's some recent news in here. So all of this pulls in from Clay, latest funding, the tech stack, what their website uses.
Jordan Wilson [00:25:40]:
So a lot of great information that it pulls in from Clay. And if I wanted to, I could do a lot more with this information. This isn't necessarily what I wanted it for. I just wanted to kind of vet these companies and these guests and to do a little bit better job of, you know, teaching me about these companies that I may not know about. Alright. So did that for all of the five most recent companies. I'm not gonna go through each of those, but great job. Right? It fired this time.
Jordan Wilson [00:26:02]:
It did it correctly. And then it says, oh, I I I said this in gamma. What did I do? Okay. Interesting. So in the prompt right. This is one of those things. I'm like, sometimes it works strangely. I ran this one as a test, I think, two or three times.
Jordan Wilson [00:26:21]:
So I specifically said to do a Canva presentation. But if you look at my connectors, I have Canva and Gamma active. So let me go down here and see what happened because if you're watching live, you actually saw Gamma pop up even though I asked it to use Canva. So interesting. So it said, now I have all the company data enriched from Clay. Let me create a Canva presentation with all of this information in a in a minimalist Apple design style. And the first thing it does is it goes into camera. Alright.
Jordan Wilson [00:26:56]:
So not sure why. But, again, you can see why I'm like, oh, when the results are good, it's great, but it's buggy and it does things right. Like, I wish I could just talk to the actual apps I wanted, and and tag them in the prompt. So we'll see if it actually figured it out in gamma or if it went to Canva. Alright. So, anyways, here we go. It did go through and give me a ranking score on kind of the fit, with our audience. So, you you know, cool cool information there.
Jordan Wilson [00:27:26]:
So made a little chart with ranking, the guest name, the company name, the score, and the why it gave them that score. So then it's okay. So what happened here is it try I told it to use Canva. It tried to use gamma. It didn't work for whatever reason. So then instead, it just made a PowerPoint. So I guess, number one, props to Claude for getting a job done even if it didn't use the right tools. But, like, what the heck happened there? Because, like I said, I ran this prop before, and it worked perfectly with Canva.
Jordan Wilson [00:27:58]:
So I'm not sure what's going on. So let me share a couple other examples here. Let's actually do one more. We'll try to do this one live. Let's see. So, let me copy and paste, my prompt here. We'll see if it works. So this one.
Jordan Wilson [00:28:15]:
I said carefully look in my sent email over the past, I didn't wanna say week, so that's not gonna work very well. Let's change that to month. Alright. Okay. So now you'll see, and I'll show this to our live stream audience here. I have a Slack channel. This is just a dummy Slack account. Now my actual one I use, just for some testing things.
Jordan Wilson [00:28:40]:
So I have Slack open in a channel. The channel's blank. So you can see this that I'd, well, if it works, I won't even have to do anything. So what I did is I said, carefully look in my sent email over the past month in my Gmail as I've sent some emails, to a colleague about newsletter feedback. Please reply back with the top five trends of feedback only. Right, then send them as a Slack message in the channel, Claude testing. Alright. I'm saying, you know, don't, you you know, reply back with this person's name.
Jordan Wilson [00:29:10]:
Alright. So, we'll give this this one, hopefully, should work fairly fastly, or or fair fairly quickly. So here, it went through. It did six steps in my email, thought, went through, synthesized some information. Right. And then it said, now I have enough data. Let me find the Claude testing channel and send the message. Again, I went through, obviously, and already authorized, Slack, to kind of have both read and write ability.
Jordan Wilson [00:29:37]:
Keep that in mind. You always gotta keep your your your data security. Right? So, you probably shouldn't just go enable this at your company unless you have permission to do so. Alright? Always keep that in mind. And then it said done. And then it gave me a link, but I'm just gonna go ahead and jump in, and there it is. Right. It is it is done.
Jordan Wilson [00:29:55]:
So, it went through, did some tasks in my email, but it took away the glue. Right? I didn't have to, you know, copy and paste something or, you know, use a different tool that's going through my email. Nope. Just did it all in Claude. It used my Gmail connector, went straight, and I authorized it to do so and sent the message straight to Slack. Didn't even show it to me. I could have had it first approve it with me, but this is just a test, and it just sent it, it just sent it right away. So let's look at one more quickly.
Jordan Wilson [00:30:28]:
So we'll jump into this one here, where I'm just saying generate a slide deck with gamma about everyday AI's unique positioning. So this is one of those instances where, it's using its, memory, of me. So I scroll through here, and it's pulling in information that I've shared with it in previous chats. You know, I have the memory option enabled, and then it's going through, and it says, I'll create a presentation highlighting everyday AI's unique market positioning. Let me generate that with Gamma now. I just got a random error even though this chat is technically old, so that's interesting. And you'll see here, it says the tool didn't return as a result. Let me try again.
Jordan Wilson [00:31:12]:
So, again, some bugs going back and forth, but eventually, it did it. So here's the weird thing. Again, I ran this exact prompt, this one, I think, four times. Sometimes I got the graphical interface of gamma to show up here. And in this case, I didn't. I had to click view your presentation and then open the link. So, it is strange, but, you know, open it up here, and it did put together a, little slide deck. Not I wouldn't use this one necessarily.
Jordan Wilson [00:31:44]:
But it actually I mean, from a information standpoint, it actually did a good job. Right? Just not really my style in terms of, design and aesthetic, but I could have, you know, prompted better for that. So, again, hit and miss. It's really good when it works, but right now, super buggy. But that doesn't mean that you shouldn't be using this. Right? I tell people all the time, this is the worst that it's gonna be now. And more than anything, these putting AI to work at Wednesdays are about you getting the reps in, for when the technology improves. Or in some use cases, it might be ready for you to go right now.
Jordan Wilson [00:32:24]:
Speaking of use cases, make sure to go repost this show. I did put together a little guide on seven different use cases using Claude's current apps that I think are really great. Right? I gave you my kind of examples how I'm using it, but, you know, people reach out sometimes. They're like, Jordan, stop, you know, showing us how you use AI for your podcast. Like, talk about other use cases, but I'm like, I I don't know. I don't I don't have live data. Right? How am I supposed to, you know, how am I supposed to show you, I don't know, these other, you know, use cases from other industries when a lot of this uses your live data. So I do have some other use cases.
Jordan Wilson [00:33:00]:
So if you want access to those, make sure to go find this, show on LinkedIn. So in the show notes, if you are listening on the podcast, we always have the link to this LinkedIn episode. Go repost this, and I will share that interactive guide going over, what I think are some seven great different use cases for using these cloud apps, just like I said. And to wrap up here on putting AI to work on Wednesday. This is the future. Is it buggy? Yes. Are there a ton of apps? No. Is it production ready to start automating a to z? Absolutely not.
Jordan Wilson [00:33:33]:
But is this something that can up your productivity today? Absolutely. Right? Especially if you are a heavy user, in some of these tools that I mentioned. Right? If you if your company uses, you know, monday.com or Asana for project management, if you use Canva and Gamma or Figma for anything on the design side. Right, if Slack, right, a a very popular right? One of the most popular communication tools in the world. So right there, maybe that's already 70% of what you do all day is you're inside those tools. So now think. You can unwind that Frankenstacking. You can start to unroll and put away that AI duct tape, all the human scaffolding that we've been doing, because this is the future.
Jordan Wilson [00:34:18]:
Right? Whether it's ChatChippy Tea apps, Claude's new interactive tools, or whatever Google's gonna roll out next. It is taking the context from the applications and the software that you already use. You are directing. You are orchestrating. But now these agentic models are sharing all that information. They're passing that context off from app one to app two to app three, like I showed you in my exam. From Gmail to Clay to PowerPoint. Right? I didn't have to do anything.
Jordan Wilson [00:34:48]:
I didn't have to connect the dot. I didn't have to poke the AI with the stick and say, good job. Keep going. Alright. So I hope if nothing else, this encourages you to start moving your day to day processes, whether it's inside Claude, Chad, GPT, Gemini, Copilot. I don't care. But you have to start doing it now because if you wait until everyone else is doing it, it's too late. And if you're an avid listener of the show, I want you to be the smartest person in AI at your company or in your department.
Jordan Wilson [00:35:15]:
So thank you for tuning in. Make sure if you haven't already, if you wanna be the smartest person in AI, go to youreverydayai.com. It takes about, six to eight minutes to read our free daily newsletter every day, so make sure you go check that out. Thanks for tuning in. Hope to see you back tomorrow and every day for more everyday AI. Thanks y'all. The risk with AI voice agents isn't that they sound too robotic for your company to use. The real risk is that they can sound too confident while saying something completely wrong to your prospective clients or customers.
Jordan Wilson [00:35:47]:
Made up refund policies, promises your company never approved, or discounts that don't even exist. You've gotta give your AI voice agents a trust layer with Modulate. Modulate monitors live voice conversations to flag abuse, false claims, fraud, and user emotions for safer, more empathetic responses. For the guardrail layer you need between your AI agents and your customers, you need Modulate at modulate.ai.