Alpha Compute CEO Explains Confidential AI and the Future of Secure Computing

[00:00:10]I'm Johnny Fernandez with Fintech TV.

[00:00:12]Now, my next guests are from a NASDAQ listed company building the future of data sovereignty. First, a best-selling author, data rights advocate, and CEO.

[00:00:22]She testified before Parliament and co-author legislation on data ownership and AI ethics. Now she's turning that experience into infrastructure. We have Britney Kaiser and the co-founder of this NASDAQ global corporation solutions now serving as executive chairman and CIO of this company Enzo Vanelli folk with Alpha Compute Corporation. Guys, welcome to Fintech TV.

[00:00:45]>> Thank you so much for having us.

[00:00:46]>> Excited to have you guys here. So, first let's get to the the question everyone's asking. What exactly does your company do? Alpha Compute is a AI infrastructure company that is focused on GPUs as a service. So we buy Nvidia chips usually through Super Micro or Dell. We get them set up in data centers and we lease the compute power. Well, we concentrate on something called confidential computing which allows all of our clients to run AI without any third parties getting access to their data.

[00:01:13]>> So Enzo, this question is for you. What made confidential computing the right bet and why build it into hardware rather than rely on software or just any regulator regulation alone? Well, the actual main reason is because we realize that about 30% of all of the the firms currently using AI are actually sharing a lot of information that's confidential and that will not be able to actually continue to go on. So we have to find ways for these firms to be able to have a secure environment in order to run these processes.

[00:01:47]>> So Britney, this question is for you.

[00:01:48]Sovereign compute as a geopolitical reality. So you call um this a geopolitical reality, not just a product. And Enzo, you've operated across traditional and crypto markets globally. So what does this mean for companies operating across the world like in Latin America, in the Middle East, and all emerging markets? Well, this is both a regulatory and compliance issue as well as, you know, a business decision. When you're thinking about data compliance across borders, a lot of times you should be running sovereign AI infrastructure in every jurisdiction where you're doing business. That data should not go into another country. It should be processed locally. Many countries have those laws in place, but sometimes that data is still being sent to other places. And then on a on the business side, when you're thinking about do you actually own your IP, your copyright, your patentable information, will that continue to hold up in court if you've been using closed source AI platforms and giving away your data. So I think a lot more enterprise and government clients of ours are starting to think, well, it it's time for us to have our own infrastructure all the way down to the hardware level.

[00:02:53]>> And Britney, you've said that the real threat is that our data uh isn't hackers, it's the platforms that we trust. uh Enzo as CIO, how does this risk translate into what institutional investors are watching out for right now?

[00:03:07]>> Well, right now there's a rush to actually build all of these actual data centers to build up this massive infrastructure, but you're building it for the hyperscalers only, which is definitely going to be feeding them and feeding their AI models. What you're not doing is you're not actually protecting your IP. So, you need to have both. And most companies are saying, "Yes, we are going to give you extra privacy, but we've seen this before. We've seen the playbook." And the problem is is that having the privacy via a contract isn't the same as as having it via a chip or locked into some type of encrypted system that is untouchable. We can't see your data when you run the AI on our compute. We can't even see your data. We never will touch it. the decentralized scalers, they have access to it still.

[00:03:56]They just tell you that they won't touch.

[00:03:58]>> So, I want to talk a little bit more about this multi-million dollar revenue contract with a leading AI research firm. Brigy, what does this deal unlock for for the company?

[00:04:08]>> For us, it turns us into a high revenue producing company. Only two months into our rebrand is Alpha Compute. So, we're really excited uh about building this bespoke cluster for a client. We've been rolling that out. That is live and that is one of four clusters that we've announced that we already have in progress. So this takes us to enterprise scale as a small and new company. And so we are very excited to be now a big player in the market.

[00:04:35]>> And Enzo, this question is for you. How has institutional appetite for this sector shifted since you first started?

[00:04:42]So, we definitely see a lot of large bets coming out like the SpaceX IPO and the commic IPO. And I think what's what's going to happen is they'll they'll be able to generate a lot of cash off of th those trades and then they'll want to invest in the next level. We are more of the edge compute.

[00:05:01]We're more confidential AI compute. So the way we think that the cycle's happening is that they're very interested in okay now that I made 10x on my SpaceX IPO. Hopefully I can now take that capital and redeploy it into to someone like us.

[00:05:17]>> Hey guys, final question for people watching around the world whether it's in New York and Dubai and S Paulo. Bring you left pol the policy world to build this from within and Enzo you've backed markets at almost every stage from the bottom up. So why this company and why now?

[00:05:34]>> Right now it is when everybody from individuals to enterprise to government are starting to integrate AI into their daily lives. Almost everything you are touching at some point AI has been a part of that process and therefore compute power is in more demands than ever before. Even you know last a few quarters ago we couldn't have even thought that the that everyone would need this amount of compute power. So we're in the hottest market in the world which is very exciting and we've been doing this now for 9 months which uh I think we are exactly placed in the right place at the right time in order to serve clients globally.

[00:06:10]>> Enzo >> no the the same I mean she she said it perfectly. So the fact is that there's a very this is a much this is a $7.2 trillion build that has to happen. Um the the hyperscalers are doing these very large projects. We're actually focused on the existing data centers.

[00:06:28]There are about 4,500 data centers just in the US alone. Um some of them have to be as you kind of you know fitted but we're we're also have the history of Bitcoin mining which allows us to have a lot of opportunities to talk to some of our Bitcoin mining partners who are now going into AI and actually put in racks inside their data center. So it's you know this is a this is like an arms race and we basically need to all be really moving ahead on it. What what we believe is that companies have to be very smart and protective of their IP and that's the the most important thing and that's why we're on the road having these conversations because whether you do it with us or or not, you should definitely be looking at what where where is your content going? Where are your most smart or trade secrets of your corporation [music] going?

[00:07:18]>> Awesome. Well, thank you guys so much for joining us here on Fintech TV.

[00:07:21]>> Great. Thank you.

[00:07:21]>> Thanks for having us.