2026-05-21-debrief

Debrief

Strategic Direction & Investment Banking Concept

• Exploring investment bank focused on helping customers hedge compute exposure

  • Leverage existing financialization infrastructure but serve as intermediary for customers who don’t know how to use it
  • AI-enabled, automated operations to handle high volume efficiently
  • Clear exit strategy: big banks (Goldman Sachs) would likely acquire specialized compute hedge boutique
  • Positions team in center of industry information flow for next decade

• Service provider approach could be stepping stone to product development

  • Generate cash flow without venture funding through strategic investors
  • Build relationships and market knowledge, then sell and reinvest in AI product

• Team strengths complement this direction well

  • Dustin: rapid pattern recognition, interconnectedness mapping, relationship building
  • Partner: technical depth, information absorption, engineering communication

PDF Technologies Partnership Opportunity

• Major breakthrough potential from recent call with Andre (CTO) and team

  • Company worth >$1B, publicly traded, sitting on massive semiconductor manufacturing data
  • Collect data from every machine/step in semiconductor manufacturing process
  • Currently underutilizing data asset due to customer restrictions (“one leak ends PDF”)

• Parametric insurance opportunity identified

  • PDF as data provider, team as MGA to underwrite policies
  • Could recreate Munich Re battery model for semiconductor manufacturing
  • Advanced packaging specifically highlighted as underserved area

• Technical integration challenge: aggregate data insights without aggregating actual data

• Next meeting scheduled, relationship ownership assigned to technical partner

Next Steps

• Dustin J Ross

  • Schedule lunch with Ronin contact for July (New York)
  • Research Sage insurance company (CEO meeting available through team connection)
  • Develop investment banking concept further on paper

• Technical Partner

  • Own PDF Technologies relationship development
  • Send thank you email requesting promised introductions
  • Research advanced packaging and semiconductor manufacturing processes
  • Write up moonshot hypotheses and strategic concepts
  • Attend Greek networking event in SF for semiconductor contacts
  • Continue building knowledge vault with meeting insights

Transcript

Them: Tea brief. We’re cooking. We’re like, we’re cooking. We’re cooking. The Ronin conversation. Very interesting. I will. I will handle that one. Yeah. That is a good relationship for us to build. They seem very smart. Very impressive. Yeah. He clearly is, like, thought about this a lot. Yes. And done something about it and hasn’t really thought about the commodification of the chip bundle, which is kind of interesting. Right. Like, but. But no, he said the memory for memory for sure. But he’s not thinking about, like. But I think logic could. It’s not a commodity, but it doesn’t matter. Like, you got. You like stocks, right. Like, like, you could still have financialization at the. Oh, okay. At the chip level. That’s something that I’ve been kind of shifting my thinking about. It’s like, you could. Again, I don’t know what that looks like, but, like, huh. That, like, planted a seed. I’m like, chips as stocks. Sure. Right. And like, you know, on the one hand, like, if. If a company only has one product and it’s one chip, then you’re just going to trade the company. Yeah. But, like, I do think, like, why can’t you trade Nvidia A200 whatever, and then also, like, trade the Sarah Brush chip? Like, you could have a whole stock market of these chips. You know what I mean? You, like, balance it for portfolio. Yeah. Interesting. Yeah. Like. Like, you could for sure. That’s. That’s like the unlock that I had yesterday. And no one’s thinking. Not no one, but, like, I have not encountered anyone that’s thinking in that way. Except for this etched guy who. He’s so busy, he’s not working on it, but he’s like. He said that totally etched is epic. Dude. Etched is sick. But anyway, I will handle that Ronin conversation. I will get a lunch on the books with him. I need to do that in New York. Yes. I’ll do it for when we’re. When we’re both there. Cool. Great. And the three of us. Yes. 100. And I. I don’t have anything on calendar for July. Like, I’m just gonna. Gonna pick it like, July. You can just start, like, throwing July 8th. Whatever. I’m just gonna make it up. So I’ll put that lunch on the calendar. That last call was sick, Nihar. Clearly, like, it went to the bat for us. Yeah. Well, also, they’re trying to please a client. Right? Yeah. So, like, that’s a very important thing. I would just, like, push you to always think about that is like, who Is your introducer. Yeah. Like, they have a vested interest to make us happy. Yep. Which is sick. And Nihar went to bat for us because you have built that relationship. And then, like, you speak engineer, like, you fucking killed that meeting. I. A lot of it went over my head also. I’m very embarrassed about this. Was distracted. Just want to take accountability for that. Not acceptable going forward. But, like, no, it happened. But I held the fort for the beginning. I understood 80 of what he was talking about. And with granola, we’ll get to speed on the road. And I was gonna say, when we get to debrief, I almost want to treat that a little bit. Like, I wasn’t on the call and have you kind of, like, run me through what your takeaways were. Yeah, not right now, but the biggest takeaway that I had from that call is, like, I do a thing, and I need to start figuring out how to do it in dj. It’s just, like, kind of difficult. Yeah. But, like, every time before I have a call, I just go to my cloud and I’m like, I have a call with this person at this company. Give me a TLDR on both. And he’s a big deal. Yeah. I. Wide. So I. So I did that. Which is why I knew I sent you the screenshot of what it produced around all their products. Yeah, I just saw that after the call. Exactly. Like. But yeah, that’s why I came in. Coming. And I’m glad I was gonna raise the parametric insurance thing. And I’m glad you did that. They are exactly the sort of data provider you need to make that. Underwrite that 100. Yeah. Like, over, like, Preston’s model he gave last week of, like, you need the metric and you need the model. Like, they are the ones that are basically in, like, integrating hundreds of different metrics coming out of the equipment manufacturers and the different nodes in the. In the process. And then they’re saying, like, we own the data. We’re sitting on it. And his answer is basically, like, we’re not really doing anything with it. Yeah. Right. So 100%. Yes. And he volunteered some big introductions. Huge. Yeah. And I want you to own this relationship. Yeah. And I want you to be respectful, but, like a dog with a butt. Like. Yeah, exactly. He offered that, like, without. Exactly. Yeah. So, like, don’t, you know, like, when you send the thank you email, be, like, also, we’d love to recap those and, like, don’t be afraid to, like, ask for those things. Oh, totally. So I think that was awesome. Well, you don’t, you don’t need to tell me that at this point. For sure, for sure. But I think the only reason why I’m saying it is like that relationship was way more value is way more valuable than I anticipated going into that call. Like I thought it was kind of just another call. And now I’m like, this could be. He could be like a critical, critical piece in terms of who we’re talking to going forward. Like, like a moonshot I thought of coming out of that is like we could partner with them as an MGA to stem. As a data provider to sell insurance. Yeah, yeah. That is a feasible business. Yeah, totally. Like, let me. Let’s get food and not talk. So this doesn’t. Okay. So I think another good next step is like I really want you to write down all these ideas that you have. These like moonshot y hypotheses. Yeah, I have an idea for something that I think we can do. It’s preliminary. I don’t want to anchor you towards this. No, it’s not anchoring. You just throw it out. I think we should start an investment bank. And I think our investment bank is focused on providing a service to customers which is helping them hedge their compute exposure. And we piggyback on what’s happening with all of like there’s so much financialization being built, but the customers don’t know how to use it. So we become the in between. It’s like an advisory as well. So it’s almost a consulting company, which is not sexy, but like it’s a service provider. What we do is we build a fucking AI enabled investment bank. We automate the shit out of everything and we start going to all these different like data center providers, data center investors, lenders, AI companies. And we specialize in helping them design using all of the infrastructure that’s being built. Because we’re a little bit late to the infrastructure parade. Right? Like ronit is amazing what they did. Like they’re 15 months and they have this exchange. Like that’s really hard to do, but we’re a little late. So what we could do is we could build like the brokerage, you know, like who uses financial markets. Like companies need to use them, they need intermediaries to do that. But we could do, and I’m not saying this is what we should do, but we could do is we could create that investment bank and there would be a few benefits. One, I’m pretty sure every big bank will want to buy that bank. So like there’s a Very clear exit, right? Like fucking Goldman Sachs. I don’t want to buy the boutique investment bank that specializes in helping customers hedge and modern computer stuff or whatever. 2. It’ll put us like right in the middle of the most interesting information flow over the next decade. Right. And so, like, I think, like, if we don’t have a moonshot, I don’t know if it’s either or what, but like, that would be an unbelievable step towards, like knowing every single person in the industry, knowing how every single company is thinking about these things. Like, and then you can build your own products. Yeah, so that’s the thing. It’s like, then you sell it and then you take all the information, the relationships that you just did and you just roll them into the next thing and like, then you build the product, right? We build an AI enabled, like, more efficient investment bank that spits off cash. We don’t have to raise venture money. Right. We just like a couple strategic investors that can connect us to place starts spitting off cash immediately. Sell it big cash payday, Take all that cash, pour it into like, AI product, whatever. So you mentioned. Oh, this isn’t a conventional product. That’s fine. Like, I’m here to like, make impact. I’m here to like, I don’t care about building an AI rapper tool, you know, Like, I agree that won’t be a trillion dollar company, right? Yeah. But it’s a stepping stone towards building a product later. Anyway, another moonshot. I’d love for you to like, develop that on paper a little bit more because I only understand like two thirds of it. But. I’m excited. I mean, I think you’re right. That meeting showed the power of like a strong, strategic, warm intro. We’re such a good balance for each other. It’s crazy because that first call was like, all finance. I was kind of just like, all right, I’m going to read on granola and ask a bunch of questions against the transcript later to ramp up. And then the second one was like bread and butter. Like, me, like, I could not have handled that call. And it’s like, you know, did you notice that you understand more about chips than Ron? It. Oh, like, that was like the D. Yeah, yeah. He was like, well, I don’t really know. That’s two weeks of work for you. Like, you are insanely good at that. I’m telling you, like, my greatest superpower is my, what I call, like, addiction to information. Like, I just sit there and like, read books, like textbooks. Like, you can ramp so fast. Yeah, I can Ramp quickly, but not nearly as fast as you. And in a very different way, I can ramp on the interconnectedness of things. That’s how my brain works. It’s because of my add, I’m very quick to pick up on, like, the. Oh, this feeds into that. Yeah, this. This is related to that. Like, I can. I can, like, that’s why, like, I can build disparate conversations on each other. Really helpful. But you can just, like, sit down, ingest, and like, it’s. Yeah. Like, my model is like, I’ll have a conversation with someone. And I’m like, I realize I don’t know that. And then like, two weeks later, I’m like, boom. Like, I’ve uploaded like two books into my brain. Like, this bank could be interesting. What? The bank could be interested. Yeah. That’s why I’m down to really start running down the financialization path. I think we’re getting a lot more kind of green lights in that area than in other spaces. Green lights. Plus, like, that’s a new way of thinking of things. Or like people who have knowledge of the industry saying, like, you know, that makes sense. I haven’t thought about it that, you know, so I get a little more excited about the financial station. Yeah. But I also know enough to know where all the shit breaks down. Yeah. That makes me nervous. Like, I think you’re probably a little. Don’t mean to project, but a little bit more excited about it because of your ignorance to it. Yeah, exactly. I’m excited to learn about. You’re like, oh, there’s. And I’m like, there’s a million ways to show. Oh, yeah. Totally. Totally. For me. For me, it’s like Denning Cougar effect. Like, yeah, let’s start a bank. You know, the bank is easier. The financial infrastructure is harder. Yeah. You know, it’s like. But the whole thing is like, where this falls apart is no one need. That’s the whole point is like, where we keep getting stuck is like, who’s your customer? Right. And if we become a consulting firm or like a service provider, an investment bank, your job is to help customers with their problems. Yeah. So you just become like knowledge factory for. Yeah. Customer problems. Right. I think we could be a pretty good knowledge factory. Like, so. And we actually, like, built an AI first investment bank from the ground up. Exactly. We could probably handle a metric shitload of work. Right. Like, we have agents that make every deck that handle every. On like, you know, and that’s where the rig becomes the product in a way. The Rig becomes what you’re. Yeah. Then. Then there’s a real unique insight that leads to when you’re a services industry. Like, you’re different. Like, the. The rig is eventually what you’re selling to people. Like our ability to run your problems in AI first way internally. You know, you’re 100, right. I think about it differently. Not. Not better. What I think about is like, your product. The product is like you and I, but we are more productive because of the. Yeah, you’re actually. But you’re actually right, which is why I’m glad you said that. It’s like you also can sell, like, higher quality, whatever. Yeah, yeah. It’s like, what? I don’t know. I don’t know, man. And I don’t want to, like, lock into that immediately because. But could be interesting, you know? Yeah, I like that guy too, Andre. I like this, like, kind of sense of humor. He was very, like. He’s very Polish. He’s brilliant. Clearly that guy is like, The company is like publicly traded, worth more than a billion dollars. Yeah. We just got a meeting with their CTO in three days. And I think the other guy’s the CEO, right? Yeah, the other John is the CEO, and then this guy Kimon, Greek name. Also. I’m going to a Greek networking event in San Francisco tonight. Imagine it might be some semi people. So I’m just gonna work the room and like, chat and I’ll acquire any contacts that are useful for us. But yeah, I don’t know. I hate service providers, but, like, it could be a really interesting time to be a service provider because I’m just like, The best thing to do is to build the market. What’s really interesting, what Ronan said is like, you studied who is the winner and all the prior multi trillion dollar. The people who are the certified exchanges. Yeah. And I was just like, that’s just a really smart thing to do. We should have done that. We definitely should have done that. But yeah, I’m just like, I don’t know. I don’t know, man. That was interesting. I just like, we talk to a company that has access to a gold mine of data, and they don’t seem to be doing anything. Yeah. Other than like, why are they only worth a billion? Other than servicing their customers with like, analytics dashboards and things like that. And then they’re like, we used to do this benchmarking, blah, blah. It seems like basically they were like, okay, the benchmarking didn’t work. We’re gonna give up on trying to like, monetize this like, data asset that we’re sitting on, right? And like, again, like, the other big screen is like, if we could convince them to be the date, like, to data provider to us, registering as a certified MGA to underwrite policies on an insurance firm’s behalf based on their data that they’re picking up from literally every single machine in the semiconductor manufacturing process. You could recreate the Munich Re, like, battery thing. There’s a query that I was doing I don’t want to dig up. Can you put that in writing? Yeah, that’s where I honestly, I just like, was not being. I mean, do you understand? Do you understand, like, what they do? I. No, and I didn’t realize what you said until today. Collect data collection from every single machine, software provider, equipment, different step of the manufacturing process. And they aggregate it into some, like, one, dashboards that like, companies can view to analyze their own yields and their own success, their own processes. But two, share the data across different companies. So in some cases they actually have visibility for like, okay, a given wafer goes through a fab and then it goes to a packaging firm where someone’s doing something with it there. And then in the fab, the equipment manufacturers, asml can have access to like, what the data the equipment is producing. And they own all that data. And they basically own that data. Why are they worth only a billion dollars? Well, because they’re under restrictions from the companies as to what they can do with that data. Right? They can’t, like, what he was saying is they can’t aggregate it. Like, remember his quote was like, one leak and that’ll be the end of PDF, remember? So, but like, if you can create more workflows that don’t rely on the aggregation. Like, for example, if you can just like, for one particular company, use all that data to understand, like, the risks to operation in a given fab and just like match that with an insurance provider to craft parametric policies. Like, they’ve already collected all the data just to imagine, like the. So that’s what they do. They just integrate data across. They’ve essentially, like. So that becomes like a technical thing, which is basically, how do you integrate the data without integrating the data? Right. Or aggregate the data without integrating the data. And also, like, how do you create like risk scoring and stuff, like models to do it based on all the data? They could be a data provider for us. That’s super cool. Like that. That’s how you break the cold start, right? What? That’s how you break the cold start. Yeah, exactly. Wow. And then they could also sell our parametric insurance policies and add on. On top of their base product. Right. Like, you know, that’s really cool if we could be an intermediate between them and the insurance company, because they probably don’t want to be in the, in the, in the, in the business of like creating policies and assessing risk. But you can match them as a data. Like that, I think could be an interesting thing to get. So, like, that, I think is like, if we take that next meeting, that is one really concrete thing to run with. Right. I think it’s a really good idea. I. I also think it’s interesting. It was like packaging. I don’t know anything about packaging, but that was the one thing. He was like, you should look at packaging. Like, there’s not enough information. I was like, so advanced packaging. He was talking about, like 3D interposer. Like, I know what all that is, so I want to see. So I was like talking to this chat. Yeah. And I was like, Okay, ready? So first I was like, Tell me about SHIFT Technologies. Remember them? That was the, the Lebanese French dude. Oh, yeah, yeah. So I was like, tell me about the company. Tell me how they’re doing. And it tells me. And it basically was like, they’re not doing that great. Right? Yeah. And then I was like. I basically was like, why is it not a rocket chair? That was the question I kept asking. Yeah. Why are they struggling? Why they’re not taking off like a rocket ship? They told me this whole thing. Then I was like, okay, do the same analysis on coalition. Yeah, that’s one of my stock phrase analogies at this point. And coalition is the answer. Then I said, What about sage? Do the same thing. I don’t really know Sage here. I don’t either. But the whole point is like, what did I ask you? But basically it was like, trick question. Actually is a rocket ship, but it just looks different. Yeah. I can get a meeting with the CEO and he’s based in Nambu. Great. Emma for my team. Her mom is the chief administrative officer there. I’m going to be in Ukraine with her mom. Can you do a little research on Shapeshire? And just because, like, you talk about the mga, but like, that’s, that’s the comp. If we don’t want to make SHIFT technologies, we don’t want to make coalition. Like, that’s now not a good comp because, like, it’s not that successful. But this is the most successful of all of them. Well, okay. Claude just like told us it wasn’t that sure. I think we need to like be a little more rigorous. Sure. Yeah, that’s a great point. I’m. I’m, I’m pretty sure they’re crushing it though. Like. Yeah, Cool. I’m gonna basically just start like keep firing off more like briefs on these things and they’re gonna get pushed to the vault and it’s gonna force you to use the vault, so. Because I love the vault. Yeah, for sure. And I’m just gonna start using the vault on my phone and we can figure out like a better interface. Yeah, I mean I could, I could like try vibe code like a app that you could just like install and open and it just like is a little friendly version of the vault. I mean that would be sick. You like for example, if you were not talking about the wall, I probably wouldn’t bring my computer. What? I probably wouldn’t bring my computer to Ukraine. Yeah. Oh, also, are you able to print Kindle or no? Oh, I’m not sure. They may, they may, they may block that based on like I don’t want you printing like entire books, you know. Okay, I can show you. You can always screenshot and print but like you won’t be able to put too much. Yeah, I just, I. I tried to download the Kindle app and like something up. Cuz I’m on my mom’s account. I don’t know. So I’m going to try to figure it out for my flight. I think we’re. I really do see the Denning Krueger effect, the Socratic effect of real life that like I know that I know nothing. Like I just feel like we’re. We knew nothing in in January event semis. Then we took. Then we like thought we knew something and then we realized we knew nothing. And now I think we’re actually really learning. It’s all because these conversations, it’s just like you can’t, you can’t learn just by like asking. You have to be talking like. And it’s like there are things that AI just can’t change and like it is the non organized knowledge that exists in like every person in the world. Yeah. All right. Anything else to debrief or you want to turn.