How I use Lindy AI to run my life

0:00I think there's a lot of people who set up things like OpenClaw, but are like, I just want to use iMessage. I just want something secure and I don't want it to do everything for me. And so there's a product called Lindy AI Assistant. Basically, anyone could have an executive assistant powered by AI that's going to be proactive in terms of reaching out to people, selling for you, drafting emails in your tone, calendaring and more. I heard about it, I invested in the company, and I had the founder come on and show us why this is an open claw killer when it comes to an executive assistant. This episode is a complete breakdown of how to use the product, why to use the product, and why I find it interesting.

0:45And I think that if I find it interesting, you might find it interesting too. Enjoy the episode.

0:53It's sipping time, baby.

1:24Okay, I'm intrigued. I'd love to learn more. Yeah. Well, we started, I think, I've been on the pod before. By the way, thank you so much for having me again. I've been on the pod before with our workflow product. And so what we learned is the main thing people did with the workflow product was building basically AI executive assistant-like workflows. Like meeting management workflows, calendar management workflows. And so we were like, why don't we just productize that and package it up in its own product that we called Lindy Assistant. So basically what Lindy Assistant is, it's an AI assistant that lives on iMessage.

2:03That's really important. Lives on iMessage, connects to your email, your calendar, your Notion, your Google Docs, all the applications you use, and starts taking work off your plate proactively. It doesn't wait until you ping it, it just proactively observes stuff and goes like, hey, opportunity for you to save time here, Greg. I think perhaps the easiest way to explain it is for me to share my screen. How's that sound? Perfect. Yeah, this is my iPhone screen. This was my day yesterday. So this is like real data. This is how Lindy starts today. And by the way, this comes out of the box, which is also another really important attribute of this new product, which is it's very opinionated.

2:45Like it's not an everything machine. It's not like a blank page where you have to figure out the workflows and the things you want it to do for you. It just comes out of the box and starts to do things for you. So morning, hey, morning, here's your Tuesday. San Francisco, it's 62 degrees Fahrenheit today. These are your meetings on your book. Inbox, I've already triaged 63 emails that arrived overnight and I have drafted four replies. This right here is one of my favorite things. Like I barely open Gmail anymore. Like it's just, sometimes the replies, I feel like I have like bad memory because I opened Gmail and I see a reply that's pre-drafted and I'm like, I don't remember drafting that.

3:23I think just because Lindy drafted it. Then she's like, yo, a few things. This happened yesterday. It was really cool. It's like, hey, your dinner tonight is at Gary Denko, but it's closed on Tuesdays. Dion moved the invite to make comic seafood, which is two minutes away. And Joshua's asking if you can meet in two hours. Your calendar is free. Do you want me to say yes? Can I just react to this real quick? When I'm reading this, it really does feel like a human being talking. Yes. We put so much attention to that. like the lowercase, the tones, right here, you can see it. It's like, I'm like, oh, fuck, yeah, please move the calendar invite if the restaurant is closed.

4:00And look, it's cracking a joke. It's like, ha-ha, yeah, we'd have sucked to show up at an empty restaurant. Sometimes it's funny because when it fucks up, it swears, you know, like it's profane. So it says, I'm like, hey, you fucked up here, I don't want you to do this. And it's like, oh, shit. It's like, oh, shit, you're right. You know, like it actually says shit, which in a funny way actually takes the edge off of the fuck up. hey, you know, I changed the location and I emailed Lucas to let him know. And by the way, the meeting is on the books with Joshua. Have a good day. So yeah, 100%, like we spent so much time and you have no idea, Greg, like it is so hard to prompt those models to adopt this tone.

4:39Like I think the voice that the models use is really basically burnt into the weight. This is why like everyone is struggling to get them to not use M-dashes. Like, look, we prompted them so much and this thing will keep using M-dashes. There is nothing you can do about this. But we've worked a lot on the prompt to make it talk like this. Later on in the day, so it prepares me for the day. That's like the daily brief. Then there's the meeting prep. So your meeting with Carnegie is in 15 minutes. It tells me who I'm meeting with. It tells me about the last meeting. It's like, hey, this is a follow-up to the meeting you had on March 4th.

5:13They are looking for an AI assistant for themselves. So the actual setup of this was what? You know what I mean? Because it seems like it knows a lot. It's two steps. It's give us your phone number and give us your Google Earth. Okay. Yeah. So it's literally two minutes and then it's out of the box. It does know a lot because what we do in the background is that we ingest your information from your inbox and from the tools that you connect, your Slack and your G Drive and all of that stuff. And then we load it into Lindy's memory. And so it actually surprises me all the time. I'm like, holy cow, how did you know that?

5:58There is more information than you think about yourself and your business and your relationship in those systems. And then little by little, it learns more and more. Because again, it reads your emails and it attends your meetings. So here, for example, I was able to ask a question about the last meeting I had with these guys. Hey, remind me what the company does again. How many folks in the sales team? They are a student connection company. They help schools recruit and retain students. And they said 50 salespeople in your meeting two weeks ago. So I can also query, it almost acts like your second brain.

6:29You can query just anything you've seen or heard, basically. Right. I mean, it kind of makes sense, right? Because your email inbox is, I mean, I have like 20 ,000 plus emails in my Gmail. So that's like so much context. I could see a world where this is also integrated into my Slack too, right? Yes, 100%. I can talk about this actually. I think it's right here. During the meetings, I talked to my Lindy. And so I'm like, I'm in a meeting. This actually happened last week. So we were in this meeting and I realized something and I was like, ah, fuck, this changes everything. We should let Ali know about this.

7:13Oh my God, this changes everything. Wait a minute, this is major because it means that Ali this whole time thought the prompt was updating, but actually it was not. Okay, Lindy, send a, like, I mentioned Ali in Slack after this meeting to let him know that we do need to force-prepare all the agents because of the eval thing. So I just came to this realization and like we were, there was this teammate who was missing in the meeting who needed to be informed. And this is the message that Lindy sent afterwards. Hey, Ali, heads up from the Reliability Sync. We discovered that the CFSAP guidelines changed.

7:46After that meeting, now we're sort of going out of order, but she's like, hey, nice call with the engineering team. I already sent that message to Ali. By the way, in the meeting I said, oh, I will provide the team with a list of top failure modes to track. And I'm like, oh yeah, can you please bring a Google Doc together and send it to Lindy, send it to the folks on Slack. and right here in Slack, created the Google Docs and sent it to the team. So when you say Lindy did this, the assistant did it, right? That's correct, yes. So yes, it is on Slack. It basically integrates, we've got hundreds and hundreds of integrations, so it basically integrates with all the tools that you can think of and that you integrate with.

8:32Cool, I definitely want to pick your brain around how to set up a Lindy Assistant so that I get the most out of it? That's a question we get a lot. Think of it as an iPhone. Setting up is something you do on your PC or it's something you do on your Android device. Maybe I'm insulting you, maybe you have an Android phone. Your iPhone just works out of the box and you don't have to worry about setting it up. It comes with all of the applications and it's got the messages and the photos app. Here it's the same. It's out of the box, it's very opinionated. We have spent so much time working so that people don't have to ask this question of how do I set it up.

9:22Let's continue here. I think this day I had yesterday was a really good example because a lot happened. At the end of the meeting it was like, hey, interesting call. I use Lindy to update my CRM. So this I did have to set up, but the way you quote set it up is like you just send a message to your Lindy and you're like, yo, after my meetings, I want you to update my CRM. That's it. And then it's going to be like, oh, what CRM do you use? Is it HubSpot? Is it Salesforce? Is it Atio? And then once you tell it, it just gives you a link to connect your CRM. End of day, it's like, hey, okay, that's all the meetings for today.

9:56While you were in meetings, you received a couple of important emails. This is an example of a time when it surprises me with everything it knows about me and my business. It's like, hey, you've received an invoice from Wilson Sonsini, that's our law firm that we work with, but they're using 1841 Market for the billing address. Is that your new address? Because I've got 101 new Montgomery on the file. And I replied to the voice memo, I do that a lot, and I'm like, oh no, that's the old address, please send them the new address. Very European of you to use a voice note, by the way. It's kind of rude, right?

10:31I hate receiving voice notes. But when it's an AI, it's fine. I send it, okay, so this one is another example of like. By the way, for clarity, I also send voice notes. I love them. And there's the transcription there anyway, so. Yeah, but it's not good enough. Apple is not doing a great job on the transcription.

10:54that's another example here of like how to set it up like i have told melinda it's literally just like a human assistant you don't set it up just tell them what you want to do and so i've told her like hey every so often i i i send you screenshots of podcasts and so what it means is i want you to go online i want you to find a podcast i want you to find the transcript of the podcast yeah we integrate with the appify so behind the scenes was it what it does is like it It finds the right Appify scraper. There's thousands of them. So it finds a YouTube scraper on Appify to get the transcripts on YouTube.

11:26And then it sends me a summary of the podcast. I feel like it's a sacrilegious thing to say to a podcaster. But there's too many good podcasts out there. I don't really have the time to listen to them. That's fair. That's totally fair. So I have, don't hate me for this, but I have a human assistant. What? I know, I know. And I hope he's not listening to this. But he focuses on three different tasks for me. One is research. So if I'm meeting someone, I want some research. If I'm interested in a particular topic, doing some research. So that's one thing. Basically, I want to figure out if Lindy Assistant could kind of encapsulate this.

12:13One is research, keeping me up to date with that research, summarizing it in a really easy way. Two is scheduling. It sounds like scheduling is built natively into this. And the third is from a sales perspective, actually. We have an agency, it's called LCA. It works with Fortune 500 companies and building AI native products, but also AI transformation. and sometimes I get tagged in like the CPO, the chief product officer of Coca-Cola comes in as a lead. I want to know about that and I also want a human to follow up almost instantaneously once that lead comes in. Yes. So yes, yes, and yes. Scheduling, I'll just start with that because that's the easiest one.

13:06Email and scheduling, that's just the job number one and two. I actually think literally yesterday, yeah, right here, right below that, I sent a screenshot to Lindy of a Partyful invite I received. It's funny because Partyful technically has a button somewhere to add to your calendar, but at this point for me it's just faster to take a screenshot and send it to my Lindy. Or I can just go like, yo, help me find half an hour with Bob. If I have access to Bob's calendar, it's going to look at common availabilities, it's going to find one, it's going to put it on a calendar. If it doesn't, it's going to look at my calendar.

13:40compile a bunch of availabilities and send it to Bob. Research, yeah, I mean 100%. So there is proactive research. So before the meetings, actually right here, this is the research I received about you this morning. It's like, hey, you are jumping on the pod with Greg in 13 minutes. Greg, CEO, late checkout studio, former advisor to Reddit and TikTok. I didn't realize that. Yeah, that's true. Look at that. I know that. And then it's like, hey, this is your still time on the pod. Thank you so much, Greg. It's an honor. Good opportunity to announce something new or drop a hook. Yeah, thank you, Lindy.

14:22I should have thought about that. So research, absolutely. And I talk to Lindy all the time. Like, oh, let me show you something good. I, well, I'm going to have to stop sharing my screen because I'm like sharing my iPhone screen right now. But I have like, I have mapped, you know, on the latest iPhones you have the action button. I've created an iOS shortcut and all it does is record my voice, send it to Lindy. That's all it does. And so I can be like, hey, in my meeting with Henry yesterday, where did he say his team was based? This is cool. And so now what it's doing is it's searching all of my meaning transcripts, all of my meaning notes, and it's going to answer the question.

15:07So I use it for research, either for information that's publicly available online or for private information, like my meanings, my Google Drive, my Notion, all of that stuff. Singapore and Hong Kong, mostly, with a few still in Shenzhen. He's trying to move them all to Singapore or Hong Kong. And that's the nifty part about integrating with iMessage, is that you automatically are integrated with the rest of the Apple ecosystem. So all I did here is this iOS shortcut. you record the audio, you send it to Lindy you open the Lindy thread in iMessage when I'm in my car I can also just talk to the car and be like, yo, send this to Lindy so all of that just works out of the box because it's just iMessage so my reaction to all this is this feels like the most human assistant I've seen yet in any product but how do you see Lindy versus Claude versus OpenClaw.

16:08Can you just walk us through how people should think about Lindy versus other products? Both pros and cons, as honest as you can be. Yeah, totally. I'll actually start with the cons. OpenClaw is a lot more powerful than Lindy. It's a lot more versatile. There's this story that the creator of OpenClaw shared where it's like, hey, it sent you to VoiceMemo. and OpenCloud went like, hey, I don't have the ability to transcribe voice memos. I'm going to build this ability for myself. And so OpenCloud is basically an agent that has access to its own code in a way and to the machine where it is currently running.

16:45So it's simultaneously very powerful and kind of dangerous because it's like an agent that's messing with its own guts all the time. The Lindy runtime is very different. It's an agent that also has a computer that is separate from itself, which is more secure, but it's less powerful because the agent doesn't have this ability to change its own code all the time. I mean, the advantage of OpenCloud is there's really two big advantages. It's more secure. Like we have been working on this for years and like OpenCloud is awesome, but it's more new. And so they are right now meeting a lot of the security issues that we've worked on for the last couple of years.

17:24And we're just way, way, way easier to use and more like out of the box. I really compare it to Linux versus macOS. By the way, it's not either or. MacOS is actually built on top of Linux. It's like a Unix kernel. But there's a reason why 2026 is the year of the Linux desktop is a meme. It's like Linux users spend their weekend installing their printer. The rest of us, we just want a computer that works so that we can get to work. So that's the same thing. It just works and it works out of the box. How do you see Lindy versus the whole Cloud ecosystem, Cloud co-work and Dispatch, I think their product's called as well?

18:11I think products are just a reflection of who they're built for. And I think Cloud is building, and this is not a job at all, on the contrary, I think they're being so successful with that strategy, but they're very much building for developers and power users. We are building for, I call him, the chief everything officer. It's the overwhelmed business owner with too many meetings and too many emails. That's the job to be done. I have to attend simultaneously to the urgent and the important. There's a thousand things flying around me in meetings, in emails, in Slack, and I must somehow return every call and take care of every customer and at the same time make progress on my strategic priorities.

18:53right so like that's really the the persona we're building for and the job to be done that will that we're trying to to solve here um concretely i think like i just said like linux versus mac os obviously cloud is not linux i do believe i do believe open cloud is linux i think cloud is more like android right it's like it is also more powerful like i have friends who have android devices it's crazy they can like ssh onto their device they can do like all sorts of crazy stuff um i think if you're like a deeply technical person who expects that out of your phone like you really want a full-blown computer in your pocket that you can do anything you want with, you should probably have an Android device and should probably use cloud, you know?

19:29Well, more like the iPhone, right? Where it's like, it's very opinionated and we pay a lot of attention to all those tiny details that culminate in the experience we're going after. And so in a way, it's kind of hard to explain, you know, like if you've tried an Android and iPhone, it's hard to point you to like a single thing that makes one better or different than the other, but it's like all of those tiny details that compound. Yeah, I think the way I see it at least is those other products are, well, OpenClaw has, you know, you have to be comfortable and super technical to use OpenClaw because of the security issues.

20:07But if you are technical, you know, it's an incredible, powerful ecosystem. The Clawed ecosystem is equally as powerful, but it's very horizontal. You can do everything, right? So you have to do a lot of setting up yourself and optimizing yourself. And some people get there, but some people don't. What it sounds like you're saying in terms of what Lindy is, Lindy is not everything to everyone. It does a few things and it does it really well. It looks at the jobs to be done of an executive assistant and it says, hey, you're a really busy person. You're a founder, you're building things, you're an executive and you need help in terms of prioritization and keeping you up to date and researching and just moving things along.

20:59And it's really just that use case. So when I was seeing the demo, that's kind of like how I was interpreting it. I'm curious where you think the next five years of Lindy is. I think what would be really cool is if you essentially look at other jobs, like a social media manager. And I download my social media manager, Lindy, a front-end engineer, a salesperson, these different jobs. And I just get a phone number and I can just use iMessage. Yeah, 100%. So I have to hold myself back from doing that every day. But that's the grand plan, right? That's where we're going over the next five years. And you can start to see it right now in behavior that organically emerges out of Lindy.

22:01So for example, just last week I was in a weekly business review and our head of customer tells us, hey, the support team is completely overloaded. We really need to open a new position. we need to find one more support person. And Lindy sent me a message after the meeting. And she's like, yo, hey, he mentioned you're looking for a new support person. Do you want me to help you look for them? And because she has all of this context about me and the company, she was in fact able to generate a list of 100 support people in the Bay Area. She understands that we work in office. And to write the outreach.

22:36She understands the company, she knows how to pitch it. So we're starting to feel that pull towards generality, but we are very deliberately saying no to the pool. We're like, hey, job one, you nailed it on the head. It's really the person who doesn't want to give his life to AI. It's actually, it really very much mirrors any technology, including the personal computer technology. The first people to use personal computers in the 80s were nerds, right? They were thinkers, right? And they are people for whom computers are their lives. and then little by little computers spread through the economy and they start to be used by people who just want to use them as a mean to an end.

23:19We have a lot of users who are real estate agents. We have a user who owns sports bars in Miami. We have a user who owns a roofing company in New Jersey. These people do not want to give their weekend to OpenCloud. They just want something that works and save their time on day one. So that's exactly what we're going after. Cool, I like it. And it's integrated. We talked about the G Suite, Slack.

23:45You said there's 100 plus integrations. Just let me know what are some of them. Yeah, I mean, everything you can think about, really, like Notion, Google Docs, HubSpot, Salesforce, ITO, Twilio, you name it. It's very rare we don't have something. It's actually funny, sometimes Lindy doesn't know herself the things she has access to. So recently there was a user, he went to Lindy and he asked her, hey, can you integrate it with my QuickBooks? And she just led him down like a wild goose chase. She was like, you know what you should do? I'm going to research the QuickBooks API, and then can you please go and create an API key?

24:21And the guy just goes for half an hour and struggles. And then she's like, never mind, I guess I'll just use my integration. So I was like, yeah, why don't you do that? So we have hundreds and hundreds of integrations. And look, if she doesn't have an integration, you can in fact just create your own API key for anything you need. and just ask you to hit up the API directly. So if Lindy's my executive assistant, I wouldn't ask my personal executive assistant to do things like, for example, vibe coding. So what are things that Lindy assistant could do and what are things that Lindy can't do and I shouldn't ask her to do?

25:01That's a great question. I would say she can vibe code because she has a computer. I'm not going to lie. Again, we are very deliberate about the use cases that we are building for and the ones we are not building for. We give it a computer so that it can sort of do anything. It can VibeCode on the computer. It is not the best VibeCoder out there, like very far from it. Like if you want to VibeCode, you should probably go to Lovable. And in fact, I think Lindy eventually should use Lovable. I think it's just going to work better. So I would say don't use Her for those very deep. like if you want like an AI for accounting to help you with your finances, you should probably use an AI that is made for that.

25:44You know, if you want this AI executive assistant, that's like your email, your meetings, updating different systems, your CRM, all of that stuff, you should definitely use it for that. And to your point, like there are executive assistant tasks that you wouldn't ask a human executive assistant. Like this, like the summarize this YouTube video, I would never ask my human executive assistant to do that. It's like, that feels like an awful thing to ask a human to do. Right. But my agent, I do feel comfortable asking him to do that. And so I think one thing that took me by surprise is initially I was like, Hey, it's not as good as a human executive assistant, because obviously like a GI is not totally here yet and all of that.