April 29, 2026

# Transcript: 2026-04-29

> 6 time blocks from 8:00 AM to 9:55 AM

---

### Zoom meeting with Michael and Todd
**8:00 AM - 8:51 AM PDT** | *meeting*

**Microphone:**
Thank you. It's going alright, how about yourself? All right. Well, I don't know if he can hear us or not or what the deal is. Hey, Michael, what's up?

That's assumed to be normal. Yes, there's a mic. It looks like he's half joined. He can communicate with us. He is. That's what he's having to do with the issues. All the AI in the world and I still can't get Verizon coverage in my house.

Hey Todd, how are you man? Still no... Yeah, he looks frozen to me. Todd can't get a connection on the internet. That's hilarious. I think we're going to have to go back to the analog data at this point. It's truly unreal, like our neighborhood. I mean, I have to go in my car to make a phone call. It's ridiculous. I think that's partly due to the elevation, but you can't get a signal anywhere. So hopefully it's good.

I don't know what happened. The internet just went. What's happening? Yeah, probably a good idea.

I've created a test script that kicks off a post request to our endpoint that can now exist in Salesforce so we can kick off individual target company syncs and then account syncs and eventually people syncs. A full-on development process set up there. That's like this morning. I kind of overwhelmed with a sales push yesterday. I kept all my messages. You already have jobs running, please wait. But so this morning, the first thing I did was run a job and I did get my single partition file. I saw what I need to see, so that is good.

Hopefully I'm going to merge this in probably live. There's one PR that's going into the dev branch. Hopefully that will be the last one that will put us in the position to use the new syncing stuff from the Rails side. So I think we can maybe start looking at how we can figure out how to test this as the first package and then maybe deploy it in one of our orgs and then maybe still continue to use the S3 buckets. Or since we use it, we could probably do our own production settings. But I think we can start looking at how we can test it, because with all the moving parts, this could be quite a test. See how that goes. But it is to the point where I can kick off tests or kick off jobs completely and I do see partitions and I do see errors when they come up. So we're not running blind. The Datadog logs in my local environment are helpful. So hopefully we see Datadog going in.

So I think all those pieces are finally coming together. I think we can maybe get this into testing a little bit more, and that would be like the first kind of release package functionality. Then we can make a decision on whether to continue down the idea of deprecating stuff in Lightning Web Components or do a kind of quiet deploy where a customer wouldn't see any changes. We're just trying to get this actually put together with everything and all the configurations good to go.

Kim, I know you always have everything else going on, but maybe at some point in the next couple of days we can talk about a strategy for that.

Over filters and changes, getting review feedback implemented, and that generally went well. In no particular order, I spent some time working on the next stretch of creating the suggested prompt filter experience on the search page for Discover. I believe I've got a separate branch that can open up and get requested changes approved, and then I can get both the SmallWorld and SmallWorld LLM side deployed.

Beyond that, I spent some time addressing a couple of customer issues that were either brought to David's attention or requested via email. Removed those users. Target companies for Brian should be able to get those imported this morning. Rally AI filters changes and any other minor changes or fixes as needed. Hopefully moving on, we'll see testing. I know there's a couple other PRs in the SmallWorld queue that we also want to get in today. Or yeah, this week, so we'll make sure to prioritize as needed. But that's me. You're up, Michael.

We'll talk about priorities, but the PRs have been sitting for days. We can't keep doing this like before where we just create lots of open PRs and then just sit and nothing gets tested and nothing gets released for a long period of time.

So let's just keep moving.

Sounds good. I had code review for Cam and just was going through some security concerns we have that we need to start being a little more careful about. Nothing new. It's all status quo things that have always been done but have also always been a potential issue.

And then I was working on something with the deal team about this morning. I have also been fielding questions from the SOC 2 team. We're waiting for a draft report. Just, unfortunately, our front line person, who isn't an auditor, isn't a CPA, but is like the team assistant, let's say. So it just ends up as a conduit for me to answer questions. Hoping we'll get a draft report in the next day or two, but it really depends on if all those questions get answered or we have a much more streamlined Linear.

And then I was also working on the settings. I'm hoping to start wrapping the first part of that project up, page by page, later this week.

Glad to see what you sent me last night, Michael, on that. Hopefully it energizes the team.

Yeah, for sure. That sounds great. We haven't caught up and had another meeting with David Connors from The Swarm, which I haven't shared with either of you two guys. So quickly on that front, because it's kind of the most important openly so far. It's three, and I would say two and three are kind of equal, right? So we've got to figure out a way to get there as fast as we can.

And that profitability can be driven by new deals and receivables. We've got great momentum with SHI. We've got something on ice with much business moving through to make sure those receivables hit. We just need more time. There's an acquisition process happening right now, which is just my guess. They've been super quiet—Anita hasn't responded to anything, Ash hasn't responded to anything. And they were going to really lean into this. Make sure we have James ready to invoice them Monday. That's for another client, and we've got things sort of back in place for them.

Cam, that was one of the things you helped me with yesterday—making sure that the analytics or the settings page was reflected correctly. We're going to have another kickoff next week.

To the degree we're getting some of these new capabilities and user experience done, and I know you guys are working on that today—Michael specifically with the Mad Lib type thing for the Prom or whatever—that would be great to have as we get the Postman thing kicked back off. But that's another receivable, a 30K to 40K deal with 30K coming if we convert it. So super important for us there.

All of them are like customers. They were open with me that they weren't going to get the volume we once were. I don't think that's the case, but maybe there is something there. So I really want to revisit that. Mike Davis is out this week. Serby's also in a weird holding pattern. They had pushed out their relaunch. I'm going to run that down with Bell. I think we've been very transparent. As of right now, they're not going to renew, so we'll be losing them in about a week and a half.

I responded with a comment that we have an active contract with them on the new business front. As I mentioned earlier, regarding how we fund this thing and what we do next, I've got a proposal for a partnership with Revo, but it would be through Zoo Ventures funding Small World. Under the guise of code development and cross-selling, there would be what we're building out and an independent investor angle, but obviously with a mind on how Revo could benefit. I'm talking to him about that.

And then there's David Connors from The Swarm. As you guys know, they're trying to get to know the landscape and figure out what they're doing—something big and fast. They're sort of sorting through it. Last week I started to realize there's something really interesting. We're uniquely different from them. They're a data play focused 100% on their data, integrating their data via their own API into everybody from Clay to Addio. They've got an MCP for Claude. It's just different. They don't have enterprise [presence]. But where can one plus one equal four? We'll just see where that goes. It's the same landscape, but they're swimming in a different sea than we are right now. This could end up benefiting everybody if we can see the synergies, and they're super clear.

So that's the big update from my end. I'm really excited about the idea of a relationship agent that's agentic. They gave us an account with verified relationship data and everything else that we can find. Third-degree relationships, ultimately helping you connect and multi-thread more effectively within your account. Looking at this on an account-by-account basis, there's great value for what our agent can do to enrich what you know about your relationships. The consumption model can be agentic, which is what people expect now.

The bad news is that still requires the salesperson to go in and trust that the agent asks correctly and does all the stuff we're imagining with the concierge. We already have that end covered.

If there are any questions for me on any of the other stuff, please ask. If not, Michael, you can maybe share a little bit more about what you've put together.

This is on top of the five years before that where we've built a lot of things trying to do kind of a new connector dashboard. The interface is also important for our users. So connectors should make it the simplest thing in the world. I don't think the complexity is the problem as much as it's the problem of choice. What do I do right now in this moment to make things better for my sales team?

One of the big problems we've talked about is that our current interface has these disjointed sections, like the relationship lead section. But it's a little distorted from the rest of the interface and it's incredibly valuable. The fact that these offer help as opposed to request introductions is good, but again it's a little unclear.

So connectors will still require getting up to speed on what's happening in all my accounts for my weekly sales kickoff or whatever. I think the concept of a deal team came from a bunch of conversations David and I have had, notes from those, and my descriptors of what this should look like. I said to Claude, use the style guide I've built for us. This is just an HTML prototype, but it's got a lot on the page and needs some paring down. It's the encapsulation of relationship leads.

It is the encapsulation of introductions and the encapsulation of matches and prospects. Concierge, year, probably. Maybe there's a deal team version that's their own thing you're a part of. And then the idea here is that this takes as much data as we can get from Salesforce—things like what stage they're at, an opportunity with a value attached to it from Salesforce. And it's matched against all of our relevance.

I described this to David yesterday as like, if you think about it for help, very strong, over 51% all the way down to these aren't even matches—these are just chances. It's like, well, this guy works or lives in one place, and here's some recommendations like, hey, you don't have any bidding sales or you don't have this or that.

But what is most important is that we enable the deal team and say, hey, your job is to get me all the meetings it's going to take to get this closed in, however that takes. And we are giving you a clear sense of direction of where it's going, getting the meetings you need. So it might be enriching, it might be requesting ratings, it might be central command for each account. This area is monthly or whatever that usage looks like. The showing goal is making it all. In fact, I would make the argument that we're going to kill this page. It's useful. Roll out the analytics with these new pages—they're built, they just need to be curated.

Here's how you need to aim, and I still have a lot more design work to do on this for a new dashboard when you know exactly what you see to see the broader things. There's some kind of kooky ideas. I was joking about this yesterday that, like, year—how you can have it send product updates or project updates. Similarly, you know, can you send my management a five-sentence blurb? There's some ideas like up. I will pause.

Yeah, I'm very excited for this work. I guess, yes, synthesize. I mean, I won't drone on, but overall, very impressed with the work here.

Was talking about the relationship leads in the introductions kind of being destroyed and bringing them into here? Does that fit? Obviously, you're not necessarily on some level. Absolutely. I mean, I think that we've maybe that are at. I'll use the word that you use—disjointed—and a unifying, all in one place to sort of look at and observe.

Todd, any thoughts or questions? It's a lot. My comment there would be, don't pay attention to the UI here, and like, because you're absolutely right, as Michael said, there's a lot of simplifying we've got to do. The way that we're looking at it as a starting point is accounts now have relationship maps with a salesperson. The agents go and say, oh, to in Hyatt how this shows up and then how she acts on it.

The one thing I do and get into this account—the probability of getting that is the highest. The second best is probably what we're calling today a relationship into that account. The underlying point here is the agent's use all of our first-party data to be your trusted relationship agent. And then you can get access to the C-suite.

And as Michael and I chatted about yesterday, we're not firming this up yet, but it's kind of less top of funnel and more mid-funnel, right? Like in the pipeline, your conversion rate is typically 25%. We're going to take that to 65%. And we're going to do it because now our job—the agent's job—is to go find these nodes of trust and to make sure that you're as wide as possible and as high as possible in these accounts.

So think of it more like we're just shifting from, because I don't—I honestly think, you know, when—and then this is where the charging for it. So our pricing model and the way we monetize this becomes very different. It's like, don't those examples that fit sort of this definition, because you're going to get offers for help and go find relationship leads and go help you understand who should be on your deal team. Who are those connectors that have done it for them?

Right, um, doing more than just asking the connectors for meetings—the agent's doing the discovery part of it, of all the data that we have would. What else to say? Yeah, I understand that. It's like, gives us a like, oh, look at all of our data. We know connectors have very specific, difficult to change those patterns. That's all just still is always concerning and like, will this break through? Perhaps I don't know. Maybe people will be impressed. A large personality because log in and then it's just, you know, that engagement is the thing that always, that we've kind of years now this.

The goal here—forget how it's presented and forget it, should be more and better and different data. And so the point is the agent, their job is actually getting the relationships together, strength rated. Their job is actually to get an offer for help that wouldn't have otherwise come. Their job is to get the meeting confirmed. Their job is to create visibility into the deal team that the rep never saw before.

And so the agent is doing and bringing new paths into view that are not visible today by the rep. Even if they knew how to use the UX that we've got today, we're actually having the agent do all of that, which is to your point, that's our issue, right? It's getting the engagement part of it.

And then the connector side, if they know that this agent's gonna do all that now—line of a corpus of data, our LinkedIn data, here's all of our stuff. Show us who knows people here. Small World does that a lot better because it's not just LinkedIn. It's looking at all this other stuff.

Michael, go ahead, sorry. When I think about why we haven't had as much engagement is going back to the relationship leads page or the introduces page. It's just not really clear data—introductions across multiple companies.

Search holistically in our platform. Yeah, making progress is the goal—that's at least my goal. There are going to be some things that they're going to have to click the button for. So accepting intros and having that stuff go live and seeing it float to the top as real opportunities—I think that's a way to drive engagement.

The other thing I'll say is the next piece I want to work on. Based on my prompting and the data I pulled together: imagine a new client who is now going after Postman and they come in for the first time. All of these fields are blank. Seeing it fill in over time—the front end with the concierge running on your goals, UI stuff—we have to pull back on that, but an agent could identify who the partners are down to the person level, and we can get their emails. I'm going to call it the relationship agent because the concierge is going to be the one that asks for meetings. So the relationship agent takes care of the rest. Just go do all the outreach, right? And we'll actually introduce ourselves like: "Hey, here's what I do to work on behalf of Postman to identify warm interactions. It looks like you have a relationship here, or we can offer help by clicking here." One click and it pops open the modal they're offering—anything else, right? And now you can even think of customers as a cool cohort for this. Like: "Hey, as a valued customer of Postman—and you know you have three relationships at XYZ—would you be willing to help make an introduction?" And then if they say no, they'll never hear from us again. Or maybe they say "tell me more" or "yes, I'd be happy to." The agent then starts learning from all the interactions and says "Okay, what's your preference of channel?" We're just waiting for warm intros through this agent, and it doesn't matter where they access it—whether through this, Salesforce, or even Slack.

We can simplify this as: here's your weekly relationship agent update from Small World on what it did while you're sleeping. It's like the transcription from Gong—it takes my calendar, takes all my emails, organizes them for me. We're hoping to do the same thing purely around relationships.

Let's keep challenging this idea. I want to add multiple paths to action on our data, which should reduce friction to a first action when logging in. I think the biggest wins will be a page where you don't need to click through. We have a lot of understanding on how to integrate this and can really be a trusted agent across the platform. And as we begin to aggregate actionable data in one place and synthesize summaries and action items based on that data—I'm excited to see what we can come up with.

I'll reiterate: we're spinning a lot of plates right now, so this doesn't go overlooked in terms of what we have in front of us—the PRs and the work on the projects in progress. This isn't to distract us; it's just an opportunity to rethink what we've been doing with a more relevant solution once we get through what we're focused on right now.

Yeah, I think there's still a lot of work to do to figure it out. To be clear, in the beginning we probably have to move very quickly at this point to prototype as opposed to design, because we can actually share those in calls and get feedback. We're pretty ready for what I call active AI. Actively AI announced a $45 million raise yesterday to do what Salesforce failed to do: build that relationship graph as we get more of this data. That's a whole other conversation about what can be monetized. People are paying for it this way.

If you want to check out one-on-one on any of this stuff, let me know. I'll probably set up some time with you guys anyway, just to catch up. We'll be in touch. Thanks, team. Thanks for being here. All right. Later.

**System Audio:**
Cam, what's up, man?

It's going all right. Hello. Zoom did its normal thing—oh, you want to join a meeting? We want to update all your software. I don't know if I'm talking first or not. Let's see how it goes. There he is. Hey, Todd. How are you, man?

Still nothing. Yeah, he looks frozen to me. But honestly, all the AI in the world and I still can't get Verizon coverage in my house. It's hilarious.

Okay, am I here now?

Yes, you are. The internet kind of went sideways.

Yeah, it's truly unreal. Our neighborhood—I mean, I have to go in my car to make a phone call. It's ridiculous. It's just bananas.

It was really strange. So I moved into the kitchen. Hopefully it's better. It's a 6 out of 10. All morning I've had no problems. And just now, of course, the meeting starts and it just goes sideways. I don't know what's happening.

---

I've created individual target company sinks and then account and eventually people sinks to use the development S3 buckets and the development AWS settings. So now we can have a sync. I've created a test script that kicks off a post request to our endpoint that can now exist in Salesforce. So we can kick off a full-on development.

This morning, the first thing I did was run a job. I got a message saying "You already have jobs running, please wait." But I did get my single partition file. I saw what I need to see. So that is good from the Rails side.

I think we can maybe start looking at how we can figure out how to test this and stuff. And then that opens the door to being able to kick a sync off from the Rails app. But that could be tested separately. So hopefully I'm going to merge this in live. There's one PR going into the dev branch. Hopefully that will be the last one that will put us in the position to use the new syncing. It has the first package and then maybe deploy it in one of our orgs and then maybe still continue to use the S3 buckets, or I guess since we use it, we could probably even do our own production.

I think we can maybe start looking at how we can test because we're not completely running blind. I am seeing the logging, the fake Datadog logs in my local Cloudflare. But it is to the point where I can kick off jobs completely. And I do see partitions and I do see errors when they come up. So hopefully we see Datadog going in. I think the things are kind of finally coming together. And I think we can maybe get this into testing in the future.

---

If it were to be a new release, it would be a quiet one where a customer wouldn't see any changes. So I don't know how we want to approach that. Kim, I know you always got everything else going on, but maybe at some point in the next couple of days, we can talk about a strategy for that. So that's it for me with that. I'll kick it over.

That would be like the first kind of release package functionality. And then we can make a decision on whether to continue down the idea of deprecating something. We can't keep doing what we did before where we just create lots of open PRs and then just kind of sit and nothing gets tested and nothing gets released for a long period of time. We'll talk about priorities, but the PRs that have been sitting for days have to take priority over the SFDC stuff.

---

I had status quo things that have always been done, but have also always been a potential issue. And then was working on concepts around security and just was going through some of the security concerns we have that we need to start being a little more careful about. Nothing new. It's all kind of a concern. We have hundreds of issues.

---

We haven't caught up, of course, and I've shared with you guys openly so far. It's like, we're going to pursue funding, number one. We're going to pursue M&A, number two, and we're going to pursue profitability, number three. And I had another meeting with David Connors from The Swarm, which I haven't shared with either of you two guys. So quickly on that front, because it's kind of the most important thing, and I would say two and three are kind of equal, right?

It was great to get the Kobe news yesterday and to get those over the goal line. It converts on Sunday. I sent Joe Tarantino a text yesterday. I said, "I assume we're going to convert. We just need more time." I have the suspicion they're going through an acquisition. So I'm scrambling to try to close as much business as possible to make sure those receivables hit. We really need to make sure we can convert these existing customers.

So clockwork: if I've got James ready to invoice them Monday, that's for another 15 or 16K. Then Postman—we got like 30 days. And obviously we got a decision process right now. That's just my guess. They've been super quiet. Anita hasn't responded to anything. Ash hasn't responded to anything, and they were gonna really lean into this.

I'm sorry. She never got back to me from my email. So I'll let you know there. But I would expect everything to pick back up early next week, as they've described to me, and we're going to have another kickoff next week. So to the degree we're getting some of these new capabilities and user experience done, that would be great to have as we sort of get the Postman thing kicked back off. But that's another receivable that we're not picking up in the activity channel like we once were. I don't think that's the case, but just maybe there is something there. But it's quiet from an activity perspective. 

As we get through some of this stuff, we're inspecting around what we can start testing to get connectors offering help. And maybe it's quiet too because there's active data.

They were open with me that they weren't going to be able to do a lot until next week once the quarter was over. So that's why we've seen a lull there. Kantar is quiet, so I think we really want to revisit that. Mike Davis is out this week. But still nothing from Splashtop despite all my outreach.

Kathleen was looking at this. Strangely, last night on LinkedIn, one of their sales leaders had posted about selling. I don't think she knew who it was. We'll see, but we're probably losing them. About a week and a half ago, he had reached out and started to realize there's some really interesting stuff. We're uniquely different from them. They're a data play now—they're focused 100% on their data and needed to do something big and fast and didn't, so they're sorting through it.

When David Connors and I spoke, and then we met in person last week, David Connors is from the Swarm. As you guys know, they're kind of a direct competitor. Part of what we want to continue to try to do is work along those lines. We had about an hour and a half together last week. I sent him a proposal on Saturday, under the guise of co-development and cross-selling. There would be a partnership where he would have a vested interest in us. This would be a partnership with Revo, but Zoo Ventures would be funding SmallWorld. They're integrating their data via their own API with everybody from Clay to Airtable. They've got this MCP for Claude—it's just a different strategy. They want to raise a sizable round of funding, as do we.

From my end, I'm really excited about the idea of a combined entity that could end up benefiting everybody if we can see the synergies, which are super clear. Their approach to solving the problem and go-to-market is different from ours. They've got a PLG model. We don't. So it's worth keeping an open mind on this. It's confidential, but honestly, we're not going up against them because they're swimming in a different sea than we are right now. The idea is to see where one plus one can equal four. It's very early, but Michael and I are going to have a call with David.

If a relationship can enrich an account with verified relationship data, our consumption model can be more outcome-based. There's a huge transition taking place right now. There's expense spending on agents that can bring more efficiency to a process. The good news is we've already built our first version of that. We have to make the life of the salesperson better. One thing we're trying to keep separate is the concierge—one being agent-led search, one being orchestration. It's not necessarily about making it easier, but making it more clear for our users. Connectors need to know what to do next.

With the new connector dashboard, we've been trying to incorporate search, and that sits on top of the five years before where we've built a lot of foundational work around relationship leads. We have built a lot of different pieces. The bad news is that still requires the salesperson to go in and trust that the agent is asking correctly, which is what we're imagining with the concierge.

Introduction is another thing—here's a giant list with not a ton of information. We've talked about having a dashboard in some form. But for AEs who want to know what's happening in all their accounts for weekly sales kickoff or whatever, I think the concept of a deal team is where we really drive that. It's a good way to encapsulate all the data we have in one place, not overly complex, but straightforward about what to do here versus there.

One of the things I've noticed is that offers for help were being positioned as opposed to requests and introductions, but it's a little unclear in this moment. Of course, I could do stuff with the dismissed and saved ones, but it's disjointed from the rest of the interface. These disjointed sections—like the relationship lead section where we know relationships are so important and we communicate that to clients—it's just a list. One of the big problems we've talked about is our current interface has home, which is to say I think this is overly complex. This is something I wanted with the concierge. There'll be data, but dealographic data, CRM data from Salesforce.

And it's matched against an opportunity with a value attached to it when the deal is targeted to close, things like that. It's not firmographic, but part of it. The idea here is that this takes as much data as we can get from Salesforce—things like what stage they're at, if there's a probity, or maybe there's a deal team version of the page, you know, a connected version of the page for deal team. We'll see. They'll still broadly use the Connect dashboard. But for AEs, it is the encapsulation of relationship leads. It is the encapsulation of introductions. It is even the encapsulation of matches and prospects directly linked to a target account. It would be owned by a request, an AE, or potentially a team. There's some thought needed there.

Build as many features into this as you want. This is just an HTML prototype, but really hit these high notes and descriptors of what I think is needed and how I think this should work. I just said, "Hey Claude, use the style guide I've built for us and just go nuts." These are things that are tangentially hanging by threads—getting them all in one place while you were sleeping. It's also illustrated up here: enable the concierge to act as your AI RDR within the deal team. And we're giving you a clear sense of direction of where it's going. And then we're just doing stuff that aligns with you.

So we lay out, much like we do in the current prospect or the current Rel.AI and concierge interface, the orchestrations. I need this sale to close. Your job is to get me all the meetings it's going to take to get this closed, in whatever form. Here are some recommendations: hey, you don't have any binding sales, or you don't have this or that. But what is most important is down here—maybe the concept of a deal team. There's interesting information throughout it. We can say, hey, here's your buying center coverage. Or they live in the same town that so-and-so works. There's a potential here, maybe not a great one, but it matches all the way down through—average rated relationships, or as I described it, all the way down to these aren't even matches. These are just chances, which is like, well, this guy had a conversation. That's a relationship lead, which is very strong. Over 51% of the time they convert to active introductions, just strong matchups reaching out to new connections. We're influencing everything. Give me one sec here.

So they know exactly what to do on their personal side. This is your task list, get it done. From here, they would then be able to go into the deal team to see the investors. When they log in this world every day, this is going to be where they land, and the deal team stuff is going to show up here too. Especially if you're on a page that looks like this, that tells you, hey, here's all the introductions and here's how you need to act upon it.

To be fair, these are essentially just the dump of all the introductions on an account so that the admins can go see. So it makes this page a little less useful. We're just charging—for every deal team you have, it's $10 a month, or whatever that usage looks like. And it leads to this, which is us potentially pricing this. This area is a little light because we still have a lot to think about, but this is where we're going to be thinking about how we're charging for this—if we're charging by token use case or if there's rating signals, et cetera. This is the central command for each account.

As we figure out broader things to do, I will pause. Cam, Todd, anything or a couple additional thoughts? On some level, what do you think about this concept of the relationship leads and the introductions kind of being destroyed and bringing them in here? Does that work for you? Obviously, you're not necessarily our talk.

This is disjointed and all over the place. You have a list of deal teams. How we set that up is something we're going to figure out, and ideas like that need a lot of thought. Some of that stuff just doesn't matter in the short term. But the idea here is this becomes the hub for how you can share it. Some product update or project updates. Similarly, you can send management a five sentence blurb. There's some ideas.

I mean, this obviously looks awesome to nerds, but does a salesperson actually want this? It's a lot. Any thoughts or questions? Yeah, it's a lot. My comment there would be, don't pay attention to dumbing it down. The underlying thing is the agents turned on for all the accounts. The agents go and say, oh, because Todd likes Slack versus email. But I know that's the only thing I've ever heard click it. The one thing I do like about this, what Michael was saying, is there are different probabilities. If you're getting notifications in the pipeline and your conversion rate is typically 25%, we're going to take that to 65%. And we're going to do it because now our job, the agent's job, is to go find all the relationship data we can get from you. But then we're going to use all of our first-party data, our agent, and our four and a half years of experience. The underlying point here is the agent's going to do it all.

And so the motivation for you now is to get your data first. We need your LinkedIn data. We need your Gmail data. We need to reconcile nodes of trust and to make sure that you're as wide as possible and as high as possible in these accounts.

Like, you know, not to use the headless term, but it's like the values and what the agent can go and do, right? Like how it's going to—and this is when we'll get into our pricing model and the way we monetize. Turn on the token wheel as soon as you click that. If the concierge set up a meeting, even better.

We've always had problems with engagement on both the requester and connector side. I mean, are connectors worse? But we've seen super requesters, so I don't know—it's just the same story that we've been fighting for a number of years now.

The goal here is to forget how it's presented and forget about them building their own thing. They're like, "Hey, here's all of our LinkedIn data. Here's all of our stuff."

When I think about why we haven't had as much engagement, it's going back to the relationship leads page or the introduces page. It's just not clear how the agent can do more than what they could do today using our platform.

So Michael, go ahead.

Sorry.

Yeah, no, that's good. I was going to say I think there are also two things to think about. It's a lot better because it's not just LinkedIn. It's looking at all this other stuff. So you're good to challenge it, and that's why we want to have these conversations.

But the hope is to require that the salesperson do it themselves, right? Which, to your point, is our issue, right? It's getting the engagement part of it.

And on the connector side, if they know that this agent is going to do all of this—and view things that are not visible today to the rep, even if they knew how to use the UX that we've got today—we're actually having the agent do all of that, which gives visibility into the deal team that the rep never saw before.

And so the agent is doing and bringing new paths. Their job is to actually get an offer for help that wouldn't have otherwise come. Their job is to get the meeting confirmed. Their job is to create—like, it's a list of a lot of potential. Is it making progress? This deal team is quite filled out.

Before your Monday morning meeting, you get ready for the week. And we've talked about how there's a very high likelihood that if that works well, then it could drive as much as it was supposed to drive in deals this quarter.

And I think that will also help to show—obviously, there's a lot of UI stuff we have to pull back on here. But just showing that you come here in an interface like this, and also having this interface be the front end to the concierge running on your goal of targeting, you know, closing the lease or [unclear], and seeing it kind of fill in over the course of the next day or two, is a good way to show the potential value. I think it's a way to drive that engagement.

The other thing I'll say is the next piece that I want to work on is important. And so giving them a clear place for that stuff to live and seeing it kind of float to the top as those real opportunities present themselves that you need to act upon.

Because there are going to be some things they are going to have to click the button for—accepting intros and proceeding with that conversation, stuff like that. They can look at their deal holistically in our platform and just go, "Is REL.AI? Is Small World working on this?"

Like, the joke about the requester dashboard is that this is not a dashboard, right? It's a search. So giving our AEs a place where multiple companies—it's a very kind of frazzled page, let's say, around where your focus is.

David and I were talking about this last week about URL-based programs. You're using Crossbeam to do this today. But that requires a lot that the rep has to do. What if the agent could identify who your partner is? I really hope this kicks into gear, Todd. That's different than what we have today.

It's like, imagine for a minute: okay, who are all your partners? You have this cross-selling. It would be great to have something like this screen here or part of the screen or something—a smaller version of this that is iframed into Salesforce or things like that.

I mean, there's a lot of stuff there. But the relationship concierge—I'm going to call it the relationship agent—because the concierge we're going to use is the one that asks for meetings. Right. So our relationship agent takes care of the rest. Just give us your companies—the names of your partners, right? And now you can even think of customers as a crazy cool cohort for this.

It's like, "Hey, as a valued customer of Small World, you know, either choose to rate the relationship or offer help by clicking here." One click, it pops open the modal they're offering.

We're driving referrals through this as much as anything else. You know, here's what I do to work on behalf of Postman to identify warm introductions. It looks like there's a relationship that you have here. And you could say, "Here are the customers that have had positive experiences"—basically a positive experience.

Learning from all of the interactions and says, "Okay, what's your preference of channel? You know, can I get better at what I'm doing?" And it's all automated, right? So like at the end of the day, the rep is just like your peer at this company. And then: "No, you'll never hear from us again," or "Maybe tell me more," or "Yes, I'd be happy to."

This agent just fucking starts waiting on warm intros through it—doesn't matter where they access it, whether through this, Salesforce, or even Slack, simplified as like, hey, here's your... I mean, all my emails. We're trying to license ourselves into a fucking crowded market.

And it's not that, right? It's like we've got an unfair competitive advantage on all the first-party data that we've got. We've got to keep challenging the idea, like poke holes in it, think about different ways we should be imagining it.

But I think if we get there, we can be perceived differently than we are today, which is a SaaS product that's like, organize it all for me. And we're hoping to do the same thing purely around relationships, right?

And so anyway, let's see our understanding on how to integrate this. And we can really be trusted.

I'll reiterate, we're spinning a lot of plates right now, and so it doesn't go overlooked in terms of what we have in front of us—just the PRs and chopping wood on the projects in view. This isn't to distract us. It's just to kind of... you know, there's no expectation to have this done, you know, in a week or something, even though it sounds like there's still a lot of work to do to figure out what it actually is.

What it ends up being is trying to do what Salesforce has failed to do with AgentForce. And so it's not relationship focused, right? So our whole thing is trust relationships, having first... I got to jump, pretty ready for a lot of calls, but Actively AI, I think, announced a $45 million raise yesterday. On an account-by-account basis, go and type at this point, as opposed to design, because you know, we can actually share those in calls and say, hey, here's what we're working towards and get feedback.

Actively AI, but ideas that don't get out there are failures by default. So anything we can do to try and test the waters. That's why I've been going straight to protocol.

I do think that, you know, as you kind of made clear in the beginning of this call, we have to move in very, very quickly at this point to get these ideas out there. And if they fail, they fail, you know, which is a whole other thing, going deep, um... anyway. Thanks for the push. It's really good.

All right. Later.

### Casual personal conversation snippet
**9:31 AM - 9:37 AM PDT** | *personal*

**Microphone:**
What are you doing? Thank you. What is it? Stop whining. Amen. I know.

---

<details>
<summary>Background Noise (4 blocks)</summary>

### Brief unintelligible mic fragment
**9:02 AM - 9:04 AM PDT** | *background-noise*

**Microphone:**
What is it like if I don't care? I'm going to go around. Hopefully they're

### Short unclear utterance
**9:25 AM - 9:25 AM PDT** | *background-noise*

**Microphone:**
Die. Let us see here.

### Single word utterance
**9:41 AM - 9:41 AM PDT** | *background-noise*

**Microphone:**
Oh.

### Brief background audio fragments
**9:48 AM - 9:55 AM PDT** | *background-noise*

**Microphone:**
Thank you. Oh. Thank you.

</details>

The April 29 morning standup with Todd, Michael, and Cam covered three intertwined threads: operational execution, revenue pressure, and product vision. Todd reported the Salesforce sync pipeline is finally testable end-to-end with partition files and Datadog logging working locally, and pushed to clear the backlog of stale PRs that have been sitting for days. On the business side, Todd outlined the company's three-track strategy (funding, M&A, profitability), shared details of a potential partnership with David Connors from The Swarm via Zoo Ventures funding, and flagged urgency around converting existing deals (Kobe on Sunday, Postman kickoff next week, Serby likely churning in ~10 days). Michael presented an HTML prototype for a "Deal Team" page concept that unifies relationship leads, introductions, matches, and prospects into a single agent-driven hub — a strategic pivot toward a "relationship agent" that proactively discovers warm paths into accounts rather than relying on manual salesperson engagement.