Time to talk about AI Nerdy topics. #AI
Chris and George B. Thomas built AI agents that pull data from HubSpot, ClickUp, email, Slack, and calendar to generate daily briefings and surface new leads. They use Obsidian with an Echo assistant to create verified timelines, and multiple autonomous agents handle everything from content drafting to team wikis, saving hours with minimal prompting.
Chapters
Good morning, LinkedIn friends. Uh, Value First Nation, welcome to another episode of Value First AI Daily, your collaborative intelligence report. It is Wednesday, March 25th, 2026. Uh, how are we doing, George B. Thomas?
I'm doing good. Um, being nerdy, uh, having fun. Seeing how far we can push some things around AI and multiple brands and timelines, which, which, by the way, on LinkedIn, the timeline post is getting more than I thought it would. I literally just threw up a picture of like, do you have these documents? As far as personal identity is timeline, bonus points for that. And I'm like, nobody's gonna care about this post. Well, I guess people care about the post and about the idea because maybe it, maybe it is important.
Yeah. Yeah.
How are you doing, buddy?
I'm, I'm doing great. Uh, had that feeling coming into today. You know, there's a lot of little things you could do. We've been showing off some pretty big things that we can build. Yeah. But when you can start to close those kind of documentation gaps at the beginning or end of the day.
Yeah.
Man, it goes a long way. And so I thought we would, we would take today to just share some of our favorite ways that we're using all of this infrastructure that we're building.
Yeah.
To just help us with the daily so that if you're out there wanting to do this stuff, which you, you almost certainly are, but you can't find the time, hopefully one of these tools or kind of methods that we're going to talk about today can, can help create the space for that time. Yeah, and I think we should just start with the one that you just, um, uh, mentioned, you know, built yesterday, um, because it's, it's a lot of kind of what, what I expected to talk about.
And, uh, yeah, yeah, so I'll just kind of explain, um, because it isn't on— it hasn't synced to this computer that I'm on yet. Um, it will, it just hasn't yet. Uh, that's how new it is. But, but let me explain. Uh, on the back of what I kind of talked about, I put a LinkedIn post out. I was like, hey, uh, timeline— by the way, Chris and I talked about the timeline yesterday of like, hey, maybe we should create this timeline where it knows George was an employee, or George was a business owner, or this is where George was working, or this is the article that George had. So that, uh, and by the way, there could be a Chris version, uh, whoever version. And so what's fun is at the end of a show we do after this show, the Customer Platform Show, I said, Chris, you want to see something cool? And I pulled up what my Echo had built while we're on a show, was a timeline of my life, historical, okay? Um, and it had these little verified tags in there where What's fun is because I'm using Obsidian and it was in Obsidian, I went for a walk later in the day and I was actually verifying things in my timeline that had been built by my AI assistant as I'm walking down the sidewalk and trying not to trip over my own feet because I'm just like enjoying it and nerdy. And it was, it was so close, but there were some things that I had to tweak along the way. And now I've got this historical timeline that I can feed it, which by the way, immediately I fed it to the mini book that, uh, we're writing together, uh, and it made these adjustments and changes. And when I re-listened to it, I was like, oh my God, that's so much better. Like, it just understood. Uh, the other thing, Chris, too, last night was I went for another walk in the evening and I was listening to the book because what I'm doing is I'm like listening and then revising, listening and revising, right? I'll know this stuff back and forth. It's almost like I'd be able to step on stage and do a presentation about the book that we're creating because we're Going through the process anyway. So at the end of the day, though, I had Obsidian up and another note, and I was giving it audio notes of like, hey, in this section of the book, we need to do this. I then turn the book back on. You know, I'm walking the whole time, and then I get to another section, I'd like, pause. Oh, in this section, we need to pop up. And so literally there was like five to six things throughout my walk that when I got back to my office, I literally said, hey, look at the entire book. Get the context you need. Now look at this URL, 'cause I was putting it in Obsidian, and look at all the things that I said we need to edit. Please create a plan to do these light edits, 'cause I don't need the whole thing to be written. And so like in real time, real edits to then give to the assistant, to walk away, by the way, and go eat dinner with my family. Come on, right? So on the back of that, late last night before I went to bed, created or had my assistant create a slash end of day because the timeline that I'm talking about is historical, but time is happening right now. And so what I wanted to do is moving forward, be able to be up to date. And so the idea is I do a lot of things like talking on this show, and I'm gonna say something that I didn't know I was gonna say, or I'll be in a meeting with a human and all of a sudden it'll turn like personal and like I'm coaching instead of just HubSpotting some stuff. And, and I want to be able to capture all that. So we created this end of day slash command that what it does is it goes out and it looks at my HubSpot, it looks at my ClickUp, it looks at my email, it looks at my Slack, it looks at my calendar, it looks at the things that we've worked together on that day, and it gives me an end of day briefing and it posts it in Obsidian or in the folder structure. Under journal for that day. And so now it has a day-by-day understanding of context and things that we talk about and things of importance. For instance, one of the things that it pulled out, uh, Chris, I've got to kind of turn over here, look at it. It even said, after I put it in the journal, all saved, here's what was written. But it even came back and it said, hey, one thing I noticed from the data that might warrant a knowledge update. Kendall and Lauren appear to be a new lead and gives me some details. And the two website migration deals suggest a new prospect. Like, yes, yes, thank you, Echo. That's exactly who those people are and exactly where they're at in the relationship with George B. Thomas, aka Psychic Strategy. Right now it goes on, right? But What I'm trying to paint the picture of is the granular detail that now I can go back and be like, imagine this, uh, can we go back 2 weeks and tell me exactly what happened on that day, what we built and who we were talking to? Like, come on. Anyway, so historical timeline, real-time timeline. Like context out the wazoo. That's where my brain's at. That's what we're kind of working on just to be able to get the best out of the AI assistant. And there has to be a conversation at some point in time, and then I'm gonna shut up, by the way, 'cause I feel like I've been talking a lot. But there has to be a conversation at some point in time with you, the human, and what you're willing to give it. Your assistant, because the amount that you're willing to give it is going to correlate to the amount that it can give you back. And there are so many people who are riding this line of like, I don't really want to feed too much personal information into it. Now, I'm not talking credit card information, it's Social Security, you know, like I'm not talking that, but like the more personal information— timelines, mindsets, beliefs, voice and tone, like values. The more you can give of that— what's happening in your day, where your brain's at, where you're thinking about going— the better it's going to be as a, like, a work partner, just like a life partner would be, or a human is what I'm saying. Okay, I'll stop there, Chris.
Yeah, it's a, it's a beautiful thing. And like, just like in the in a CRM when everybody's inputting data and like going through the process and you get to the next meeting and everybody trusts the data is in the system, you're not having to confirm what happened since last meeting. Like you just come ready every day.
Yeah.
Right. The kinds of conversations and the kinds of value you can create because everybody's always ready is what you can kind of, you know, correlate this with.
Yeah.
And like the power of that is like, yeah, you can do context engineering and you can make memories and you can make like, hey, AI, make sure you're doing this and this and this. But I've gone away from like it's less automation and it's more autonomous activity built into the skills, right? And your end of day, I'm going to share my screen here because a lot of this is related to what we're trying to do with the collective. Uh, like teaching people how to share screens, not the screen that they're looking at, but from StreamYard itself, you know. Um, and like the infrastructure we've been talking about, all those agents, like what are they? Why do you need so many agents, Chris? Well, when I do things like, and like mine's called Daily Recap. It's, it's the agent going to look at HubSpot. It's an agent going to look at all the shows that ran. It's an agent going to look at my calendar. It's— and then those agents are designed to do things with that information, like update the repo, like update HubSpot, like create the report so that I can see it and I don't have to look at it in the terminal, right?
Yeah.
And suggest, hey, you had this call. Right. This, you don't have to add insight. You have to draw it out.
Right.
Ryan Ginsburg and I were talking about an upcoming, you know, client, um, you know, scoping call number 2 from the Value First Scoping. And we were talking about, you know, the, the feeling like you need to add value in these calls, like by bringing expertise and like, and it's like, nope. That's not, that's not what we're here to do. We're here to ask questions and draw the insight out of them because they have it right. And we actually can't give it to them about their business and what they're going through. Right. 3 hours later, after the call with Ryan, which that transcript just automatically flows into the, into the repo.
Yeah.
And at the end of the daily recap skill, I get a suggested article based on some call from that day.
Yep.
Right. And this was the one. And I was like, yep, nailed it. Show it to Ryan. He loves it. Right. Those are the kinds of things. And like, as I'm having that meeting, I'm not thinking about creating an article based on what we're talking about.
Like, no, God no, right?
Like, sometimes those ideas, but like having AI here, and the cool thing is like maybe I'm not ready to publish it right now. It's, it's there, it's got the idea, and it saved the idea, and it's got a list of ideas now to, to, to be able to publish and know where to go to get the original information. And that's just with one skill, right?
Yeah.
There's a skill at the beginning of the day, daily ops, that goes and looks at the calendar. Which meetings do you want to prep for? And all of these things are how we use— and just to— when we talk about the kind of infrastructure you can't build alone, But when, when George and Nico and I like talk about it, we've been, I'd say we're power users of this stuff at the very least. We go pretty hard on this stuff. And what we have built for ourselves has not happened overnight.
No.
It's also much easier when all you have is yourself to worry about whether the system works. Like for you.
Without a doubt.
Across a whole team, right? So part of the interesting challenge of the collective is like, how do you make this kind of stuff accessible and not have a loss of like experience and quality, you know, as other people come in, they're gonna use it in their own ways, right? Yeah. Like if we can do that, that's, um, yeah, so we've been working on that and that's part of, that's the primary value of like this when we talk about AI infrastructure. It is also like the workflows and the skills and the like, you know, slash daily ops slash end of day, like getting access to that stuff. Right? Like you could come over here and I've been trying to find it on the website. Uh, and there's some downloadable skills, right? That you can, but if you're not putting it in the right spot, if it doesn't match with the agents that you've got, like in the memories that are happening, like in your structure, right? It can be hard. It's hard to get it right. Um, and so like shout out to, to Aaron Wiggers, uh, who has started building this collective platform, which is basically like Slack plus agents, so that all of the 70 agents that are in, in the, you know, infrastructure that I've been talking about every day, making it so that we can interact with them, like, so V, Pax, Sage over here, like direct messages, uh, also like interacting with our, with our humans, the other humans, right? Getting guidance. And while I would love for everybody to just get over it and get into the command line, like, not everybody's doing that.
Yeah.
Or do they want to? So this is, you know, I'm optimistic that this kind of, um, moment where you can just come in and say, hey, what are the 12 complexity traps? Right? What does this mean? Like, help me out with this deal, right? What's going on in HubSpot? So very similarly, like, it's connected to all the systems. It's got the skills and the agents underneath it. And now we can also allow for learning. In the system in a way that I'm not one to worry about things like— I'd rather push instead of like govern. Yeah, right. Like, let's see what we can learn, right? But the ability to have, you know, adding one more human being, what I, what I found to these systems, it's not even close to 2x, it's like 10x at least. Yeah. If it's somebody who wants to go as hard as you, it's like 100x real quick.
Yes.
Right. Yeah. And so, but it only gets there like if, if the person doesn't have to start by building like their own infrastructure. Right.
Yeah. It's interesting that you're bringing this up because I think it's, it's even if the human doesn't have to build their own infrastructure, I would even say it's like if AI doesn't have to build its own infrastructure. Let me tell a story to explain what I mean. Yesterday— no, 2 days ago, I had a call with a human who wants to migrate from WordPress to HubSpot. But in the conversation, they also realized and we realized that they need a fractional HubSpot super admin. And I said, well, there's 2 ways of doing this. I could give you a quote for just us migrating the website, doing a light design refresh, Or I could give you a quote where we're your HubSpot super admin and we also are doing the migration for you as your super admin, but also then being able to do forms and workflows and other things and, uh, kind of flex or flow with the priorities of what you're trying to do over the next 5 to 6 months. She said, send me both of those, please. Both of those. And by the way, it doesn't have to be an either or. It could be like a, yep, I want that and this, like I want them both, but But she wanted both quotes. And I want everybody to know that you can buy back your day. You can do things in a way that it never has been done before. Because what I did was give the transcript of the meeting to Echo and said, hey, create these two quotes. I need titles, descriptions, terms. I need a comment to the buyer. And I gave it some context. Here's the terms I'm thinking on this, here's the terms I'm thinking on that, make sure that we say this in the comment, and then boom, it built them. And then I did some, a little bit of manual stuff in HubSpot, sent an email a day later like I said I would with 2 quotes. But the time invested was minimal. By the way, you heard me talk about taking 2 walks yesterday. The time was minimal. Buying back time in your day from the work to do the things that you need or the things that you love, right? I need to get out and walk more. I'm just saying, I'm being honest for a second. But here's the thing, during that prompting of the story that I just told, I said, hey, make sure you're a sales rep and HubSpot expert. And I started thinking about all the work that AI had to do to go figure out how to be an AI expert around quotes. And I thought, you know what, I'm a helpful human. How about I help my AI assistant truly be amazing at HubSpot? To which then we went and had it just grab all of HubSpot's English knowledge articles and bring it into the folder structure, therefore bringing it into Obsidian. Therefore, when I say be a HubSpot expert, it says I already am. Because it has the information, it has the context. And also, on a side note to this, now instead of me spending more time going to Google, looking for the knowledge article, copying it, pasting in email, sending to another human, I now have my own internal HubSpot wiki in Obsidian where I search for the term or word and it will pull up the articles, air quotes, information around the thing. And now I can say, hey, use this MD file, create me an email with the instructions of what the human needs to do, and can you go ahead and create that as a draft in my Gmail so I can go over and preview it, add to it, and hit the send button, please? Buy back your day, ladies and gentlemen. Figure out how to get your assistant to truly be Smart, heavy lifting, like amplifying who you are, human-powered, AI-assisted. Yeah.
Yeah. And if you are like, because we get it, like if you're on a team and like a lot and there's teams around you, there's people that are going to be impacted. Or you have to go through, you know, various decision makers to confirm you're allowed to do things like this. I think every organization, there's probably an opportunity to make internal, like, internal enablement sites better, internal resources better, right? Because one of the things that I do is like at the end of the week, you know, there's a weekly recap.
Yeah.
And from that, there is a build log that gets posted on the website. And like when I think about the experience of Value First and I treat it like a product, Like one of the hardest parts has always been letting people know what I'm up to.
Yeah.
Like what's new, right? Yeah. And having that motion of just like, hey, hey, hey, B, you know what we did this week? Like go ahead and just throw an article up there. And then you start to think about it like, oh, software companies, this is what they do. Like they have a build blog or like, here's 2.1, here's 2.2, here's, you know, all the changes being made. And when you can do stuff like that internally, like just so you don't have to say, hey guys, we just made this update and here's the, here's the one-off email that we're sending about it. So if you didn't catch it today, hopefully you can find it in your email inbox next week. Right. And then I'm not going to get mad at you when you didn't see the update and you'd have no idea what's going on. Right. Like all of these communication moments, when you can see this from that perspective, it's like the documentation of knowledge as you go.
Wow. It's fun to listen to you talk right now, Chris, because I remember it might have been 30 days ago, a place that you are at, a conversation that we are having. And I think it's valuable to share it to just in general, uh, humans who are going through getting ready to start whatever this, this AI assistant process. To hear you talk about there's a beginning of a day, there's the end of the day, there's a weekly, there's a— you, you took these words that I said that I want everybody to realize are maybe some of the most powerful things that you can do from a mindset standpoint, and you, and you executed it. And that is, Chris, do you remember when I said, bro, just one brick at a time? One brick at a time, dude. That's all you got to do. And I just heard in that last piece that you're talking about, uh, this brick, that brick, give another 2 bricks, maybe 3 bricks, and, and over time, one brick at a time, ladies and gentlemen, That's when you start to build the special things that you hear us talking about, one brick at a time.
Yeah, we're going to keep sharing those bricks. By the way, not always sunshine and rainbows. I, you know, when you get to building all this stuff with technology that you're not used to using, I just hit a limit in my sanity plan. So that's why I'm getting, uh, stuff on the website that I don't like. So that's something I gotta handle. But guess what? Uh, it's been the free plan so far, and I feel like I've gotten enough value out of using Sanity to happily upgrade.
Uh, yeah, especially since their pricing isn't that, um, yeah, it's a decent pricing. I've been, yeah, I've been paying since the beginning. I was like, that's a no-brainer.
So again, yeah, catch us tomorrow. We'll be talking about more bricks to put in, in your stack, in your day, to help you do more with AI. Thanks so much, George.
