#8 Alexander Oelling (Founder & CEO, INXM) on Building Deterministic AI for Enterprise Processes Reliably

So far the narrative was clear you had always to spend more tokens to get more intelligence and hopefully it will bring you to more reliability I think from the test bed we've created that's not true. You can have more reliability in spending less to zero tokens This is Cutting Edge AI brought to you by Angel Invest with your hosts Jens Lipinski and Robin Harbord Our guest today is Alexander Oling founder and CEO of NXM and they are creating a reliable execution layer for business processes. At the core of NXM is its AI process execution engine The system which orchestrates AI to automate and execute complex enterprise workflows. In this episode we will explore a new approach to AI that Alex and his teams are pioneering in Instead of relying on expensive reasoning at every step the system of NXM initially uses a large amount of computing tokens to learn how a process should be executed. Once mastered that knowledge is compiled into a plan and can be executed much simpler and more efficient. The result is process execution that can be up to 100 times cheaper, more reliable and fully audible for enterprises. NXM just emerged from stealth with a 5.7 million euro preceded round by Jerry Ventures, Redstone, Angel Invest, Linden Capital and others. Alex has so much more to talk about. The good thing is he also has a podcast that compiled AI podcast so if you want to know more about Alex as a person and what NXM is doing I can really recommend the podcast it's launching soon. But today this is the cutting HCI podcast by Angel Invest. Let's go. Hello Alex, welcome at cutting HCI. Thank you. Yeah, great to be here. And also hello to Jens. Likewise, what's happy to be here. So Alex, you have been a founder already several times, but what are you doing right now exactly? We created NXM and there's an interesting story around that since we spend the last years in the aerospace industry building rocket launch software, electric helicopter remote control software and so on and EAP systems and those kind of things. We figured out that we need better AI. We need more secure AI, more transparent AI. Yeah, and so higher the practicality of the AI integration and we created NXM to address all of those issues. For those who don't know you can you quickly describe what kind of hardware did you build over the last like five years before you started in the next. Yeah, actually we were always building the software for the hardware as you know, modern complex products always involve software to get functional and that's what we did. So I was previously at either aerospace building or helping to build the spectrum rocket and the launch system around that and my team was responsible for building the launch control software and the, yeah, the entire software you need on a shop floor to produce actually the rocket. You whether have in house developments or external software that you have to integrate and roll out over the organization and before that I was part of Volocopter building an electric helicopter, EV toll, which was later than sold to a Chinese company. I was helping do it there, building the full integration suite called Volo IQ, which is a kind of an ERP system and an end to end from booking solution for the end customer to book flights via an electric helicopter to remote control of the aircraft, airspace integration, crew management, all those kind of things. And yeah, so it was also a huge project with a couple of hundred developers involved and it was free AI. So that was one of the last really big software implementation projects in my opinion. I think area of those projects you just mentioned would deserve an own podcast, but we want to focus today on in XM and what are you building in AI next time. But what happened at which moment did you realize, oh, I want to found my own company. I want to focus on AI right now. What happened? I already created two companies before. The first company was focusing on next generation database systems. The second company was focusing on property technology, so door lock systems and for industrial areas and those kinds of things. And after that, I joined ISA Volocopter. And I think for me, it was clear that I wanted to create one day another company. And as I saw that there was actually a gap of mature AI companies that actually deliver and as a curing sound solutions for the enterprise sector. It was clear that some of the developments we did in house in the AI sector were pretty useful also for other companies for other manufacturing companies also for banking and finance sectors and those kinds of things where you have a lot of data, a lot of unstructured data that you need to structure. It was also clear that a pure LLM is not enough in order to ship an enterprise use case properly, so secure that if you can repeat them, the process in the enterprise and get always the same results. And we call this determinism. And when you have an SAP transaction, for example, about 10 million euros, you send from A to B, you rely actually on the perfect outcome. So any mistake costs you 10 million. And that's actually where the gap is of the adoption and the prices. Maybe one simple question in between what does in XM stands for because you described this compile AI just just right now, but where's the connection to an exam? Yeah, so in XM stands for in X machines or intelligence X Machina, we played around with it. And actually it's pretty hard to find a short new company name with AI because a lot of AI URLs are already gone. Our idea here is that we deliver a system that brings back the control into the computer of the user. So it's not just in the cloud or a solution. It's actually running on your own hardware if you want in your own stack. And in X machines stands for in X machines. So it runs in X machines. And what we've created here is actually not just a wrapper around another LLM. We created a new form of AI that is another way of doing reliable AI transactions without the well known ways of reinforcement learning, for example, or other learning methodologies for language models. In XM is obviously its enterprise ready AI system. Now there are a couple of fundamental insights that you've had. But you said, look, we can't just take an LLM. We have to and put it into enterprise and let it run loose because there are a number of real problems with that. And the problems that as far as I understand them are number one, if the LLM just goes there, you know, it's intelligent, it hallucinates, it makes things up. It's not repeatable. It's always different minor differences, but it's still different. And when it's always running, it's also expensive, meaning you always have to do a, you know, burn tokens and the tokens cost money and eventually they think was so much money that that you can't do it. And then the fourth problem is that you're always pulling back to Redmond or to wherever the service are. And it's also other issues around that because of knows where your data goes and you really don't know what's happening. So all sorts of problems that we say, well, I would really love to adopt the AI, but I've got problems left right and center and I don't know how to get them done. Can you just explain to the people are listening how exactly in XM is fixing that. And because I think what you're doing is like a paradigm shift in AI for enterprise. So if you explain how it works so that they can they can start to understand how significantly different it is, what is super useful. Yeah. So at the first point here that you actually mentioned this reliability. So you need reliable AI transactions. We've created a way in getting or forcing an AI to create repeatable results over time and we call it compiled AI. We are actually tracing the path of a result vector of the result vectors of your answer and making actually another form of AI that we call compiled AI that is able to, yeah, to reproduce certain results. For example, you have a transaction in SAP or in another enterprise system and you want to repeat that transaction. You actually don't want to have hallucinations, but you want to have a mathematical correct way. One set here's a new invoice book that invoice through the ERP system. And in order to get there, you need you can use an AI you can use several agents, AI agents, each agent is done creating a lot of tools. and the results might be not correct over time. That means you will have hallucinations in each process step. And what you want is you want it actually in AI engine that once you told the engine how the correct way looks like, it does always the same result for you. And tracks on through the system, clicks through the system always the same path. And we found another way of doing that, which is kind of a secret source on our side, but we actually created an agent system that is able to create pre-compired, compiled AI subvector space and put it into an actually executable as you would normally execute on a computer processor. And that actually transaction, or this actually workflow takes 80 to 100 times less tokens for repeatable transactions, then compared to use another agent for each process step. And that's actually something you can quote 'cause we're working on really seeing some benchmarks for that, also with some well-known players in the industry. 80 acts tokens on repeatable transactions and enterprise use cases and enterprise processes is actually more than a game changer. It changes the fundamentals of the industry, where right now tokens is margin, is revenue. That's what Nvidia set on stage in Taiwan, right? So right now everyone thinks, I get better results if I spend more tokens. But the reality here is there are technologies out there and we have one of them developed and invented in a form of more or less, we think we have created more or less an AI frontier lab for applied AI. And we've created a new way with compiled AI to get there. And I had an example for that, right? Imagine you have a robot on the floor and like your vacuum robot, right? In the past you always had to repair this vacuum robot, right? Then agents came and you could say, please repair it for me, AI agents, right? Please repair it for me and the robot will stand up again when it's falling down. But if you transact that the next time if it's falling down, you actually spend the same amount of tokens again. But the path is of success for getting the robot on again, online again, might be the same all the time. You just turn it around, put it on its feet and it starts walking or driving again. And that's what we call compiled AI. All of that success vectors in the LLM get be compiled in another executable. And that's what we put in our runtime environment on the robot. And yeah, in the future the robot not even needs to have internet connection in order to get on the feeds again, right? Because, and we're not spending tokens on that, right? And I think that's a very important fact here. So basically you take the LLM, the LLM does something complex and you say, yeah, that actually worked. Yeah, that was a good way of doing that. And this thing gets compiled into something that's really, really simple, that runs over and over and over again. That's 80 to 100 times cheaper. It's repeatable, therefore it's auditable. You can also run it in your own personal environment. You don't have to put it onto the outside. You can also share it inside the organization, say, look, 10,000 people, here is the way of doing that. It's always the same. This is the way that has been approved internally or whatever it is. This is the way to do it. This has been certified, security, safe, reliable and so forth. And then effectively, what you then do is you have-- Yeah. --created a really, really dumb mini AI system. Even the mini AI system. It's like an artifact, if you like. That's correct. And from the AI, it's like-- I think I'm actually actually the right word. So it's really kind of eligible. And it just knows how to do one thing, which is to this thing. Pretty much. Which actually is not some people. And we have some early customers, right? And there was a procurement guy that asked, like, oh, I don't get it. It's actually the same like an enterprise process automation. He said, yeah, the difference is you actually prompt your process automation, right? And it's clever. It heals itself if it's got broken, right? It has also a ton of functionality you will never get with them in other ways, right? Or with an enterprise process automation. And it's actually super fine granular. And it's kind of scaled. So we had a conversation, a couple of months ago, where you told me-- where you got the point and said, oh, wow, you don't need any consultants anymore to set up an SAP migration, for example, or that kind of things, right? Because you actually take things that usually take months for implementation, or usually in the enterprise, you connect a lot of enterprise systems together. And you need a lot of consultants in order to do that. And now you can just have one guy that can prompt something. And actually, it works, right? You can create an enterprise transaction within seconds instead of minutes or weeks or months, which usually is attached to 100,000 of your resources meant for external consultants. And that's actually not just a shift for the consulting industry, but for the entire-- yeah, AI industry, since so far the narrative was clear, you had always to spend more tokens to get more intelligence. And hopefully, it will bring you to more reliability. And I think from the test bed we've created, that's not true. You can have more reliability in spending less to zero tokens, actually. And that's kind of cool, right? Yeah, I think it's like a total paradigm shift. Because you're basically saying you, for the inventives, trying to figure out how to do it, or if you then it stops working, curing that, you need to spend money. But then repeating it is almost free. I have almost-- And the vast majority of the cost is effectively-- it's in the same way which train an AI system, and then you have inference, and you just ask it. In your case, the inference, and doing it all over again, repeatability that becomes-- which is the vast majority of the cost that becomes almost free. I have more things. I said, sometimes I have to feeling we've created an AI laboratory, a frontier lab for applied AI. Since we've learned so much from our own experience, but also talking to enterprise customers, there will be more forms of AI and also the features and functionality. Let me describe a bit the components you need, or which we released. First component is, of course, something we call orchestrator. It's a highly sophisticated AI model and combined, actually, with an agent. And that's actually the AI you are talking to. So that's actually where you prompt your stuff, where you prompt your enterprise processes. That's where the creativity happens. That's not so special. It's just in LLM, fine, you want to end the price processes, and a lot of data that we had from previous projects, but also that are available through partnerships, through some of the larger companies in the industry, that we also, in the background, did. That's one component. The second component in that is, actually, we call something that is the compiled AI itself. We call plans. And a plan is actually a very highly sophisticated JSON document, which can be very large, 100,000 of project steps. We did that because we wanted something that we can audit, that we can trace, that we can, as a human understand. It's too big, usually, to create it manually by hand, by someone. So you need the agent to create it, because otherwise it's too complex. That's where, actually, LLM are really good at right now. On the third component, we have the execution engine. So we call the entire new sector that involves here enterprise process execution. So we all know our PAs, robot process automations, we all know ERP systems, enterprise process, tools, and so on. And I think that enterprise process execution is actually that what happens. It's actually a new way where AI actually does the work. And that's the overall meme here. So actually AI that finishes the job. And that means you have 100,000 of invoices per month. You need to process market research. You have procurement deals, tons of that stuff, that need actually human manpower, but also where human manpower lacks this understanding, high complex environments, and relationships in data. So you can create here workflows and plans that then you can even give a way to someone else. A plan actually is not that much different to like a file format. Like it's an executable, but it's an artifact. It's like a PDF that you can send someone that does the invoicing and demands and planning on the taxation side, for example. And if you execute it in the execution engine that we are delivering, It's actually running in the context of the user. So you have have actually work thanks to AI, thanks to compile AI, that you can actually deliver to someone else even. So it's portable. It's insane. There's a ton of that technology stuff that we've developed in the past years. And we're happy to share that stuff very soon. I think it sounds to me as if, if you think about it, what that means is that, at least partially means is that there are the work that humans do is now in a JSON file. Yeah, I mean, the JSON file is the person who does that's effectively the description of the work that is being done. Look, we as humans have a small mind. We have a small memory, right? We can only remember three to seven things at the same time. It's really hard. If you have a thousand or more, sometimes 10,000, gyro tickets, zero tickets at the end of the month to kind of close a certain milestone in an engineering company, for example, you have to decide what of those tickets you can put in the backlog, which was already fulfilled, how are the requirements on each ticket? It's a huge job. You never have enough people to do that. I have, I mean, right now, maybe 20 different use cases of things like that, where not enough employees, humans actually can put their brain in in order to solve a problem in time before a certain deadline. And that is in procurement. That's in engineering, that's in sales. If you have a public offering, which you want to attend, and you only have 10 days time to deliver a certain offer, for example, you have to, you know, scroll through thousands of pages of requirements. There are a lot of problems, which are right now larger than a team of people can solve in a certain amount of time. And that's actually the sweet spot for technology like that in the first place. But I think it has the potential in the long run to transform the enterprise at all. Since once you have technology like that, you can have different processes. It also has an impact, of course, on the workforce, on the mid to long run. But you can't do all of that without the practicality, reliability, transparency of that technology. That was actually the, the, yeah, the optimization curve that we've been on since a while now. It's not about creating the best AI. It's about creating the best reliable AI and the best dynamic harness system around it in order to really fit into enterprise use cases. Can you give us a couple of use cases that you can talk about, probably? I think the use case actually everyone has is invoice processing. It sounds super stupid. There are tons of solutions out there. But you know, what, once the AI over big aerospace company said to me, Hey, Alex, what should I tell my internal people when you get over 95% accuracy on this invoice processing stuff? While the best technologies I can buy in the market, and we can build ourself above, above 60%. And we have a million invoices per month that we are processing in hundreds of people that do manual work on them to put them into SAP and to, into the ERP systems. And I think it's a super easy use case everyone has. And it's much better resolvable with compile AI. Why? So first of all, traditional machine learning only underscores to understand the invoice. That technology here combines and splits the large use case of understanding that invoice into several atomic processing steps, each sandboxed in and part of the compiled AI process step means first of all, you have to understand the invoice. There might be 400 different versions of options of that in certain invoice. If you invoice number can be put it on different areas and the invoice, whatever, stack in the text, understanding the OCR text is a complex thing at all. So there are good LLMs out there that can do that job very well. But then you usually get not above 60% accuracy. What you really want to do is understanding the entire process behind of it, every SAP transaction, every requirements document, every order number, every orders in the past, comparable invoices on each invoice. And you wanted to use not one AI system LLM to do that. You wanted to put each processing step into a separated, secure, transparent sandbox and make sure that there's no hallucination around that. So means that there's only a limited result vectors acceptable. And that's actually what compiled AI does. It divides the big problem into thousands of little steps individually and executes them very well through. Therefore, you come over this higher currency at the end because now you have literally a whole LLM step and coded in thousands in two thousand simple steps that only have, you can force that, those simple steps into result vectors. So at the end of the day, you get a perfect result because it's not just the invoice, it's also every transaction around that invoice that's understood and can be put into context, including the right transaction in the ERP system that it belongs to. And if you think about that, normally you need a certified SAP specialist to put that invoice into SAP. It takes 20 to 30 minutes per invoice. Think about that. How much time you spend on that stuff. And if there's a new invoice that no one saw before, the system will reason and find a way to solve it. And if it's a certain likelihood, it will process it automatically or it's under the likelihood. It will escalate it to a user that then does the manual job. But this doesn't stop here because now you've trained literally the AI on solving that kind of invoices again automatically. And the whole ontology creation, which was previously in those other systems, a manual job is done automatically, for example. So there's a lot of things here. And it's actually a very complex prompt to get this working. But that's only one time. And it's literally done in half an hour. And it's only one use case, right? So let's just assume that there are quite a few of those kind of use cases and companies. And then eventually they will all get resolved. And then a lot of the manual work shifting bits from one screen to the other, from one application to the other will get automated. If we lift our gaze up beyond that multi tens of billions of problems and say what's actually behind that? And how do you think about that? I think the first time I heard the phrase autonomous company was in 2017 or something like that, when people started talking about it with the machine learning applications that were around at the time. And it felt like that it might potentially be achievable to get there. But it obviously was a far more garner that wasn't that wasn't doable. Not so with the LLM's same story. But how do you think about beyond solving these kind of use cases? What does the future look like from from where you're setting? I think that you we have to talk about the time frame, right? Right now there's still a lot of politics in larger companies, right? There are still separate areas of the company which are little kingdoms, right? Where people don't talk to each other properly. Decision processes are very slow. So in the past, you've tend to pay people by their experience, right? And the longer someone is in a larger company, the more usually the important the guy is or the girl is. And now, and that's actually from talks with some chief people officers on very large automotive companies and in the past months, you pay people for the outcome. It's an outcome based payment. Means you pay people for two things. You pay people for asking the right questions and delivering the high quality outcomes. And in between, there's AI and AI agents that you don't have hundreds of agents. You have a couple of specialized agents using enterprise plans with a technology like compile AI that do reliable and produce the domestic results. Which repeatable results. You pay someone for still being in charge that those transactions have been done correctly. So asking the right question means creating the right workflows that are running in the context of a certain user that's responsible for it. And second, controlling those results is actually the task that someone has now in a future corporation. And there might be special roles for that. I call it on the third pillar invisible AI. Since this kind of those kind of systems are not delivered to 100,000 of people inside a large company. Most of the processes like this invoice use case as stupid as it sounds. It's a complex one, but there's not that much people anymore that actually see that process. It's kind of indistinguishable for magic as Arsacic Clark says. So when technology is fine enough developed, you can't separate it for magic. So it runs in the background. And I think it's not the God AI that knows everything and can do everything. Here, it's more the Star Trek vision, right? Where you have an AI that runs in the background, is always there for you and does things automatically, which we don't have to care anymore about, since it's reliable, it's a repeatable, reliable and to be honest, stupid process and boring process, like an inverse processing that's then done automatically. But it can be also backlog solving in Gira for milestones. There are so many things, right? Improculements, supply and negotiations, pre-automized things like that that are now with enterprise process execution. Yeah, possible. Every time you speak about a specific process, like with a ERP system or in Gira, you're always working with the systems which are already in place. And so far as I've understood, you don't want to replace those systems, right? Yeah. If I assume this right, why did you do the decision to do it in that way? I think that there is an enterprise systems of record. And that's actually the systems that are actually there, like the ERP systems, BLM systems, MES systems, and so on in production. And then there's the systems of action on top, which we see right now. Right? So that enterprise process execution is actually a systems of action. And it takes actually actions in the systems of record. I think that there is probably 100,000 of microprocesses we call plans to be implemented first before you can exchange the systems of record layer underneath. So if you have reliable transactions and let's say you have the digital twin of your organization in plans in compile AI, then you can start to exchange ERP systems underneath. Right? Because then you're literally interfacing to those systems through the systems of action, through prompt, through, okay, let me release one other thing we've created. It's called a Generative UI and the deterministic Generative UI. So one of those plans can create you the interface itself on the fly. And it's deterministic. It puts exactly the right fields in the system underneath under the ERP system. So you are interfacing and getting away as a user from those kinds of systems of record. So you probably have never to log into ERP again. So at one point you have to digital twin running of anti organization, then you can start to exchange the systems underneath. And it's then more or less a data platform underneath needed. Some of the enterprise we're talking to have already created those data platforms. And we'll be so happy soon when I can share the use cases and what we're doing there. It's so cool. I mean, some of them really did a great job in creating those data platforms over the past years. And now they are really kind of learning the gains. So they've done the right job, the right decisions, and now they can implement those kinds of AI on top of it and be very successful. So then you're saying in the long term, a company could run on the compiled AI of an example to store the records and do actions. I think you never do an innovation alone right there. I might be other people doing the heavenly same idea at the same time. But compiled AI is actually not our term. It's the a term of some researchers in Stanford. There's a paper around that the fun fact here was that we've created started not to writing papers. We've created a system a while ago and then figured out that there might be some other people working on the same things. So I assume there will be more people doing that. And enterprise plus execution is definitely another category here. By the way, fun fact, how we get recognized of those. We have 24/7 running agent with plans that those plans are understanding every website, every news article, every competitor and put them into a structured output. We have collected more than 180 gigabytes of pure data and it's not MD files, right? So it's executed in compiled AI and it runs 24/7 costs a couple of 100 euros per month. It actually runs through the entire web and it found in a report compiled AI understood what it does and recommended us to have a look into it because it looks similar to that stuff that we have in our documentation and in our code base without that we recognize it ever. Yeah, you can build something like that and it's by the way was just a prompt with two sentences. We typically ask what's cutting edge about what the company is really doing in the most cutting edge. I think in your case it's like it's like a combo of cut after cut after cut. Yeah. So when we first talked in April last year. Yeah. There was no, I mean, this was a few months before you even started the company. Yeah. How has your vision since then changed? Like what was this? What was it? What were the like? Did you have all of those thoughts already at that point? Yeah. How did they come over the course of the next three to six months or nine months? How did this actually evolve? Of course, most of the credits go to Matthias RCTO because he's a genius. He worked on that stuff since years and the team around him. We have a world class team which could be also at meta or at at anthropic or open AI and those kind of guys and people have so much experience over the years and also some great researchers. Most of the stuff is on their mind, right? So I was talking to them a lot and understanding the things. But when we met, I think I understood that there will be an AI orchestration. So there will, there's an orchestration layer that kind of can steer enterprise processes. And that's when Matthias ideas with compile AI mixed up with, I think, my ideas for the enterprise orchestration. So funny that we created, you know, the pitch decks from last summer, right? And then you see, for example, SAP stepping on the set fire conference on stage and announcing the agentech AI enterprise, right? Enterprise AI and it's so funny that you, you know, have pitch decks that are more than a year old having the same claims in there, right? And I think it's always, you never have a loan or certain idea. There's always like other companies, other people doing the ideas at the end, execution. I think you came up initially when I was in the audience, we not had created a company and you had a presentation, right? At this conference, you had some slides in there where you were talking about the enterprise layers, right? The UI layer in the systems of record underneath, like all the other systems, we said that there will be players on the interface side. And this, you know, enterprise software layer gets thinner and thinner because it's getting eaten up from the AI systems underneath, right? And that matched 100% to that what we think as well. I think over time, and you see that now with Salesforce that are going agentech and changing their entire product, SARS offering in that direction. So they're getting smaller and smaller and the new interface actually is prompting its agents. You see that also on the laptops now with the sandboxing stuff that Microsoft released on Windows a couple of days ago. So you get your agent and there's just the fundamental gap. And I'm not sure why no one wants to really hear that or wants to accept that, that spending more tokens will not bring us to more intelligent systems at the end of the day. We are reaching, we're reaching kind of plateau, even on traffic was releasing that in a block article today or yesterday. And they understood that as well, everyone understands that. But the enterprise itself won't adopt those technologies if they are not as reliable, as mathematical possible. And you only get there if you force clever AI and it's math at the end of the day. It's vector operations that you do. And there's a mathematical solution which we found in order to do that problem. And I think it's actually logical if you think about it that the asymptotic that we've now reached a point where the AI, if you spend more tokens, cannot improve the end result. And this is completely logical because the AI is right now, probably at the point where in these kind of environments, it acts at the human performance level. Yeah, absolutely. So therefore it was 10 times better than the human, but it does a human process. How can it deliver a better result? It can't because it's just more expensive, right? The system that's been designed for the humans effectively cats the ability of the AI to perform because it can't outperform in that environment, in that context. It's just not possible. Yeah. So therefore, what you need to do is say, okay, we're now at this point. So now what we need to do is basically optimize what is happening, improve the speed, improve the reliability, transparency, auditability, security, or whatever, all these other elements, the cost more than anything else, but spending more money on that. that will not make it any better. And I think this is definitely true for the face of the next few years where effectively processes done by workloads done by humans will get partially replaced by AI. - Yeah. - But then the question is then when it's transcends, when it gets to the next level up or five levels up or whatever, and then these systems are no longer made for humans, but they're made for the AI. And at that point, the performance comes into play and I think it's just very, very hard to envision what that would actually look like in real life, right? Because we haven't got anything like that right now. - Yeah. I think that we have to accept that there's a European approach here, right? And no one stops us in doing so, right? Just because of the entire industry things, more tokens are better. Doesn't mean you can think the other way around. And there is always a technology that makes something more efficient. Why did we stop in developing technologies that makes things more efficient? That doesn't make sense, right? So the approach here is use fewer tokens, have more rules, have also things like on-prem functionality as a runnate and use it for the actual production manufacturing use cases, make the deterministic control and more practical, reliable, auditable. And by the way, when the cloud is the end word of everything, and we are creating here in the cloud next-god, it gets faster and faster and faster, why can I create with some open source models from China, actually on my 10k, 10,000-year hardware stuff here at home, right? On-prem, right? We specialize AI hardware. Why can I have similar results like a year ago on the cloud? Because it gets more efficient. And there's not the one answer to that, which means the cloud will produce God at the end, or the best AI or AGI, name it as you want. But the thing here is, we can have some of the stuff pre-produced in executables, in plans. You have a lot of storage that doesn't-- that stuff does not consume a lot of storage. In four terabytes, I can put 1,000, 100,000 of their plans. And you use your local agent that runs locally on your machine, using that plans as it uses tools and databases right now. To execute things. So it can be everything from reading your email to calendar, to more complex things, which involve more people, put other colleagues in the loop, human in the loop, decisions in the loop, use other enterprise systems, and so on. There's nothing that stops a certain plan and doing complex things, right? And it's self-healable if there's a problem that will heat itself through the agent around. So whatever, it can run on-prem, because you don't need unlimited tokens in order to do a certain process reliable. And that's actually maybe also the reason why NVIDIA, for example, putting a lot of effort in new processors and stuff that run on-prem, that run in a laptop, that run in a workstation, and so on. There's more out there than the stuff that you can actually see. And I'm really happy to share now. We weren't stuff mode for a couple of months, right? I'm really happy, actually, that we are now able to share that technology. And over the couple of next months, we release also more and more of that stuff. There's also some PhD guys on our side that are working on new papers and so on. So it's really cool. And I have to thank you guys for the trust, right, that you've put into us in order to make this a reality. Yeah, I'm so happy. [LAUGHTER] Of course. [MUSIC PLAYING] Thanks for being here. If you enjoyed this episode, support us, believe me, follow and share the cutting-edge AI podcast. See you next time. [MUSIC PLAYING]