Daniele: If we are talking about digital, it's not installation of toys, it's a transformation of the business. You need to look at the business side, you need to look at the technology side, and at the people side. These are really the three transformational elements. And very often, companies were just focusing on technology. Announcer: You're listening to Augmented Ops, where manufacturing meets innovation. We highlight the transformative ideas and technologies shaping the front lines of operations, helping you stay ahead of the curve in the rapidly evolving world of industrial tech. Here's your host, Natan Linder. CEO and co founder of Tulip, the frontline operations platform. Natan: Good morning, Daniel. Welcome. Daniele: Thank you, Natan. Thank you for the invite. Looking forward. Natan: Yeah. So this is a very special episode. You've been running around Boston for the past week. What brings you to town? Daniele: Yeah, so my, my peer from farm operations, the head of quality and the head of manufacturing, we were with the leadership team of the head of manufacturing, just visiting Boston and actually also looking at companies in the digital space to exchange learnings and understand how to best scale digital across our industry. Farm operations. That's Natan: amazing. That's like a digital transformation class trip. Daniele: Yeah. Correct. Correct. Part of it. Yes. And Natan: now we're sitting and we're very excited to have you hosted here in our production studio at Tulip HQ. So great to have you. And is it the title for this week for you like Digital Transformation Reloaded? Daniele: Yeah, I think so. I think so. It's a good word. I mean, it's really about not only technology, right? We are also looking in, uh, how do we accelerate the delivery of the business goals that we have? How do we, let's say, manage the complexity that we are facing? Also looking forward 2025, 26, we see a lot of complexity coming in, which is positive on the one hand, but also really requires us to think about new ways of working. Natan: Yeah, totally. And you know, I like this word reloaded. You know, the matrix, right? So it was like the first one, the second one. And then I think there was a pause and then it's like, okay, the core concepts are so good, we don't need to make matrix number seven. We just need to reload it. And I kind of feel the same with digital transformation because it's kind of like since, I don't know when exactly, 1995 and on, I guess, PC emergence with an okay operating system, then we never actually stopped digitally transforming. When, when would we'll ever be over? Daniele: You know, for us, we don't believe that it's only a project or a program, right? Again, for us, we are not looking at installing toys. We are looking at transforming our business. How you work. Yeah, the way that we work, the way that we develop our medical products, the way that we manufacture and supply them, the way we are, let's say, ensuring highest quality standards. We want to do it in a different way. Why? Because we see that, uh, let's say our product portfolio is changing significantly. We are launching more products. We have a higher product mix. 60 percent of our portfolio will be delivered through devices. I think this new changing reality, we will not be able to manage without making sure that we are leveraging technology without making sure that we are building new capabilities. We having a software solution, which are helping the business to really work differently and generate value. Natan: Yeah. So I think this is a perfect segue to understand your role because you joined 2022, if I recall. Almost three years ago. Almost three years ago. It's amazing how time flies. I always found The way you define the role is super interesting because it's marrying the digital data analytics with operational excellence. And of course, operational excellence have been very traditional, rooted in the factories, in the shop floor, in the sites where, where the work is actually being done. And speak about this, this combination. It's super interesting. And how are you approaching it when you're thinking through transformation? How are you bringing all these functions together under one roof? Daniele: In my previous role, when I was at McKinsey, I was serving many, many pharmaceutical medtech companies. I've seen that, as you said, digital was separated from operational excellence. That's also tradition. It's, it's. That was traditionally. It was traditionally done like this. Separates them. And also separated from IT. And many times digital was an IT element. And I think that's fundamentally not the ideal setup. And I see that many companies are actually shifting. You're Natan: being diplomatic about it. Exactly. Like in the day and age we live in, it's like a recipe for disaster. Daniele: And that's why I think that my role was set up right. I mean, myself, I'm part of the business. I'm part of the pharma operations leadership team. My peers are the head of manufacturing network, the head of quality, the head of regulatory, the head of supply chain. I see my role part of working with the business to generate value. And I think that's a fundamental relevance. If we're talking about digital, it's not installation of toys. It's a transformation of the business. You need to look at the business side. You need to look at the technology side and at the people side. These are really the three transformational elements. And very often, in the past at least, companies were just focusing on technology without understanding actually what kind of business value you want to generate. And that's why my role really inherits these three elements, business technology and people. And we are combining operational excellence, redesigning business processes, making sure that we are working in a different way, making sure that we are looking at the user at the center to understand what the user needs, so that we are designing AI, let's say capabilities, digital tools, which fulfills these user requirements. And that's really the integration, which I think is key. To really maximize value in digital. Natan: So I want to dive into that, but maybe let's ground a little bit because of course, Roche is known as a global life science leader. Can you give a quick overview? What is most exciting now in the product line and where the technologies in, in what you're delivering to your customers? So what is Roche focused on nowadays? Daniele: Roche, we have obviously a diagnostic business, somatic business, and we have a pharma business. I'm working in the pharma operations business. If we're looking at our value chain of pharma technical operations, we have technical development, manufacturing, quality, supply chain, regulatory, sterling gene therapy. It's quite a spectrum that we need to look at when we are talking about digital and advanced analytics, for instance. And in pharma, we are looking at delivering transformative medicine to our patients, right? We have five therapeutic areas, for instance, oncology, neurology, cardiovascular, so really transformative medicine, which we believe are fundamental to, yeah, save patients life. Natan: Yeah. So, of course, and highly regulated. Daniele: Highly regulated, yes, and very complex biological processes. Natan: And huge cost on R& D. High Daniele: investment in R& D that we need to do in order to make sure that we are always, let's say, innovative and delivering really, yeah, as again, transformative medicine which are delivering value to Natan: our patients. You know, going from a clinical pipeline to operation, it's what we're hearing in the domain is the pressure to get NPI down, you know, new product introduction down to the factories, the pressure to understand how quickly you can ramp up operation, the pressure to, you know, with all the mRNA revolution, like actually going into low volume, high mix therapy of one, the. Are you seeing all that playing into how you're thinking about your digital transformation strategy? Daniele: Yes. If I'm looking at the focus area from an operational perspective and a business perspective, I mean, one fundamental element is we need to be faster in the way we are developing our Medicine, right? Again, we are launching more and more products. We have a higher product mix. Natan: That's super difficult with regulatory, like Daniele: saying Natan: that we want to be faster. Daniele: Faster in developing and faster in supplying our products to patients. So end to end lead time, development, let's say time. It's what is key, one area. Second area, again, Products are very complex. Our processes are very complex. We need to ensure highest standards in terms of quality, robustness, and compliance, which is also a fundamental. And the third aspect is overall efficiencies. As you said, we have a few areas like cell and gene therapy, but also other areas where we see that COGS plays more and more role. Right? So being able to drive and manage complexity at lowest cost to ensure that we also providing medicine at lowest cost to society, which is one of our key purposes, right? It's fundamental, and it's so complex that it's not easy to drive obviously efficiencies and still making sure that we are providing the most complex and transformative medicine. And that's why we believe that, I mean, we, without, let's say, big data analytics, for instance, AI, really digital, we would not be able to really manage this complexity. Being faster, at highest, let's say, quality robustness, and driving efficiency down. Natan: These are all very clear. Catalyzers, drivers that push the, you know, from the current state to, is it almost like, can you think about it through the lens of you're creating a new production system for the company, a new, you know, production system are also very rooted in lean and operational excellence. Daniele: Absolutely. From two dimension, right? If I'm looking at the production system, one, let's say renewal need is that we're seeing lean production system is not anymore. What we need, we need a kind of integrated production system from digital and operational excellence integrated. Natan: That's super interesting because you have been practicing lean as far as. Absolutely, absolutely. And that has been awesome. For many Daniele: years, for decades, right? Natan: That's a very sort of established mindset and all the thinking and operational practices in the company. But you're saying something very fundamental now. Yes, it's important. It's okay, but it's not enough. So I'm not going to get to the next level. Daniele: Absolutely. Again, because the complexity of our portfolio, for instance, of our processes is really requiring us to create a paradigm shifts in a way that we really, let's say, leverage new technologies. So one element is really, when we're talking about production system is the integration of digital and operational excellence to create a new production system. Natan: So humans and data. Daniele: Humans and data, right, making sure that we can take data driven decisions, that we are more intelligent, more automated, more flexible, modular, all these kind of elements. The second aspect is a bit also, I would say the topics that we are looking at them, right, where we maybe in the past, we're looking at Changeover optimization. It's still one element, but we are looking at also more complex things. What is an Natan: example of a changeover? It's very clear to everybody. Exactly. It's very basic. Changeover, we're thinking about, okay, we need to take this production set of assets and not make drug A, make drug B, but this relatively similar equipment. Let's measure time and safety and deviation. That's clear, changeover. What is an example of a more complex process you're looking to implement? First, Daniele: I would say that we are not looking at single and isolated use cases. I think that's an important aspect. We are looking at capabilities that we want to build and value pools. And a value pool is a cluster of use cases, which help us then work differently. So for instance, Product health. Now products are biological, high variability, high complexity, multivariate influencing factors are driving actually the stability of a product. So we need to actually pull down different use cases together to be able, for instance, to predict cell aging, to improve yield and titer, very complex, again, multi influencing factors, which are driving those kinds of things. And that's where we are applying data analytics. Second aspect is that we believe human and digital solution need to interact. In our operation system, we have multiple systems, a very fragmented landscape. So we are looking at solutions like no code, low code platform, Tulip, for instance. How do we A, integrate all these kind of systems together? So that we have a quarterback system, which is helping us to be more efficient. Natan: I love a quarterback system. I don't, no one called us that yet, but we're going to steal that one quarterback system. It's Daniele: my IP, my IP. Yeah. But I think that's really required, right? Also to reduce complexity of users dealing with multiple systems. Yeah. Which creates complexity. The digital Natan: overhead. Exactly. Often I get this, uh, you know, in conversation, like people spend a lot of time in thinking through their new architecture for digital and their production system, but then it's there. And like the net benefit is like what digital overhead. So like even too much digital is sometime an issue. Daniele: Yes. And the second aspect is really also scale. Scale, I mean, a bit of a buzzword, democratization of digital, right, or data. And this means for us, I mean, our population with over 12, 000 people in just only in pharma technical operations, not all of them will be. and should be data scientists, right? So we need to make sure that the full population is able to leverage data, digital analytics. So no code, low code platform is a great way in actually expanding the ability, the capabilities to all our colleagues so that we are making sure that they are solving their problems by themselves. Right. Instead of pushing solutions, we are allowing them to actually create their own solution very quickly in a very modular way. So that really they are solving their problems because they know best. Natan: Yeah, they have the context of the work. You know, when you're talking about scale, of course, In many of our conversations, this term scale comes in because everybody wants, by the way, I don't think it's like wrong. I'm just saying, you know, they have this aspiration for scale. Yes. But then always this one word comes in and this is the adoption. Yeah. And I think that's the thing you're talking about. And the comments you made earlier, like from, you know, looking at capabilities, and it's so interesting the way you describe it, because you didn't say a word about I don't know, architecture or the buzzword, it's so interesting because it's almost a definition for what is composability. In the organizational sense. Yes. I want to talk a little bit about what supports this kind of transformation, like from an architecture perspective, but it's pretty cool to see that the business strategy is starting from where, where will we make impact? That's what it sounds like. Where would impact be? And then I will try and scale that up. But I agree, like the democratization adoption is like always been and will be a challenge, I think. Yeah, yeah. Daniele: And again, I mean, as I said, in the past, I've seen many companies really primarily focusing on technology. What kind of platform do we need to use? What kind of infrastructure? Rather than first discussing what kind of value do when I generate, what kind of capabilities do I need to build? And what is the underlying infrastructure that I need to Provide to actually fulfill these needs and knowing that we are working in a very dynamic environment. And this means, as you said, the composability of the infrastructure of our architecture of solution. It's fundamental for me. It's really about, I think, monolithic solutions, right? Obviously creating rigidity. Which we want to really overcome because we see that our portfolio is dynamic. We see that our needs are changing very quickly, also depending on external influence factors. And that's why the composability, the modularity of our solutions are fundamental. Natan: So for people who are not as familiar, I mean, you've been deep in, you know, composable architectures and refining how we actually Turn a technological strategy that comes second, actually, to business impact. Help people understand in your own words, like, what does composability actually mean? Like from the organization vector and from the architecture vector. Daniele: I mean, for us, composability means that we have the ability to create modular applications, which can then create an ecosystem of solutions. Right? Why? Because it gives us speed in the way that we can start actually creating value. Right? We are trying not only to maximize value, but actually to also increase the speed to value. And this composability, modularity help us to start with maybe a smaller area, where we're starting to prove the value. We increase, let's say, the adoption, because people are getting motivated. They see already that their work is differently. And we are creating benefit for the day to day work. And then we can extend actually the ecosystem with additional modules. Natan: One thing I'll add to that, that I've been hearing more and more is composability allows us to stop predicting the future fully, because if you know the future, exactly, you should totally do a monolith and be done with it and move on with your life. But the reality. is that you can't because, like you said, there's business pressures and then, okay, you have to speed up there and you have a quality issue here and you need to train a thousand more people. Daniele: Yeah. And it's really from a business perspective as we, right, we were discussing this dynamics that are coming in and more resilience, but also from a technology perspective, right? Do we know what kind of technology in five years is the best one? We don't know. Why should we now spend hundreds of millions? Yeah. Into a monolithic solution. Natan: And by the time you implement go live and we know like over whatever it is, I think the McKinsey numbers are 70 percent or more of these projects are getting delayed and over budget over time and so on, by then architecture is changed because technology is not static, right? That's why I like when people talk about their architecture, kind of like bring them to first principles, you know, like, um, I'll talk about it through, you know, my engineering and architecture lens, but, you know, what made the World Wide Web work is, uh, hypertext protocols that give us a good addressability, you know, sitting on a good, uh, seven layer communication model, and everybody said, yeah. You know, hypertext and linking of pages, that's, that's a good way. Let's continue with that and like build on that. And that was like engineering principles that were set up in the way the internet was architected, if you like, that survive until this day. And I argue that they will survive later. And of course, that's a technological perspective. But how do you think about The first principles when you are thinking about your new tech stack, when you're coming, okay, I'm helping the organization because I'm sure you have a lot of voices. The operation wants X and IT is saying Y and leadership wants Z. You're in the middle, right? Daniele: Yeah, I mean, as I said, I'm serving the business. All my peers from quality manufacturing, so on, so in gene therapy, but I'm serving also and working very closely with IT. So I'm really making sure that we are combining the two. I mean, if we're looking at overall, I would not only say. Architecture decisions, but I would say as a whole, let's say software decision, right? We have a clear make or buy type of strategy and we are looking at elements obviously like total cost of ownerships, obviously, right? We are looking at also speed to market. How easy, how fast can we actually introduce this software solution? How easy is it to integrate to our architecture? Are we then So we're stuck with this specific, let's say, system or partner, how much flexibility do we have, right? All these kind of things. But that the first is we are looking at what are the user needs based on the user needs. We're defining technical requirements from an architecture perspective, from a functionality as a whole perspective, and then it's pure make or buy. There is also a strategic element that we're seeing. And do we believe That's rather a commodity capability or a strategic capability. For instance, if we're saying we believe we should own the capability of artificial intelligence, which is driving our product health, which understand our product behavior in detail, that's a strategic capability we would not outsource or buy a software solution. Right, we want to build our own capabilities. But for instance, the way that we interact and integrate human with digital solutions, for instance what Tulip does, broadening really the skill set of our people and making sure that they can really create their own apps. That's for us a capability which we believe we should not own. There is something in the market which fulfills our user requirements, which is fast in terms of speed to market, which is modular and flexible, which can connect to our architecture. That's an important element. Can augment it. Exactly. Connect and augment, right? We have obviously MES system, LIMS system. UMS, the whole stack. Exactly. And again, having really a quarterback system, which helping us to connect and augment those for our user. It's fundamental. Natan: You said something that really stuck with me, that understanding the requirements of the users, which in my head goes all the way down in the size of your enterprise to multiple sites at the end of the day, the specific quality lab next to the line, maybe the packing area or how do you do this requirement collection at the size of the enterprise? That's what Roche represents. That's, I think, something very interesting to people who are, you know, grappling with this because I think at the project level, okay, we all understand we, we lived and breathed Agile for the past, whatever, 20 years, but you're seeing it like for such a different sort of scale again, help people understand how you, how you're directing the size of organization to understand user requirement at that level. Daniele: Yeah, as I said, we are grouping and clustering use cases to so called broader capabilities of value pools, which are already helps to scope and prioritize things. Then I think it's an important element is involving users early on with so called local product owners. Yeah. At the site level? At domain level, I would say. Right? It can be a site, can be a, a lab. What's an example Natan: of domain? Domain would be quality or domain would be like, um, Daniele: can Natan: be a quality Daniele: specifically in our, I know, raw material lab or whatever can be packaging a filling, let's say, uh, area depends. Concrete Natan: operational areas that are spanning the, the network Daniele: where we wanna really transform the way that we work. We have lo, so-called local product owners. Mm-Hmm. , they come with really the. Process knowledge and a functional knowledge. Then next to the local product owner, we have so called global product owners. They are looking at from a technical perspective and ensuring that the requirements of the different local product owners still allows scalability across the network, right? So that we are not building, let's say, again, monolithic or isolated solution, point Natan: solutions. Yeah. Daniele: Yeah. That's right. And we are collecting those requirements from a local product owner perspective as well as from the global product owner so that we are ensuring tailoring to the local needs, but allowing global network wide scalability. Natan: So I have a follow up question on this that looks at it from top down, bottom up. So the tension comes because continuous improvement teaches you that if you're closest to the problem, you should kind of listen to what's happening in the, in the field, because there's a lot of empirical evidence for that. Of course, that if you eliminate that waste and do the net value closest to the issue, you create the most value to the customer. The question I have on that It's like, how do you reconcile? Because it almost sounds like you need to have an exception handling mechanism in your push from top down, right? The top down is what gives you adoption, scale, pacing, speed. And I think this is the crux of this new production system that we're talking about. Daniele: Yes. I mean, first of all, I would say that Roche is a very empowering, decentralized organization. So, we obviously also are making sure that the bottom up approach is lift, right? There's a culture that listens and innovative, very people oriented, very empowering so that people can building their own solution to solve their problem. But as you said, at a certain point, if you want to scale, there needs to be also a central element into that. There needs to be a balance between bottom up approach, empowering, making sure you're building the capabilities. Peace. That's early adoption and motivation, activation of the organization, and a more standardized way to deliver full value at scale. That's the combination. And what we have done is, for instance, from a governance perspective, we have created a tech council. And this tech council has basically a couple of responsibilities. One element is an advisory body saying, Hey, if you bottom up right from the shop floor coming up with this solution, this idea, we advise you on what kind of technology decision, what kind of standards you should So that we know that you are building something which will be scaled also across the network. That's one element, tech advisory and standards guidelines providing. Second element is obviously funding, where we see that there is acceleration required. Natan: So funding would come from your body centrally to say, I see something happening in the site level that could be impactful across the whole network. You want to amplify it. Daniele: Correct. There might be cases where the business is immediately saying, Hey, I'm funding it because we need it. Exactly. We need it. And I have to, there might be either, or again, let's say situation we were seeing, We want to accelerate and we are from a tech council, we are funding it to really make sure that we are creating the right capabilities with the right standards. So I think that's the really a government body which helps to, again, allow a bottom up approach, but still creating some central standards, some operating model, which helps to really ensure scalability. Natan: So those are ways you're getting. Knowledge sharing, which is critical because I think one of the flaws of the old production system, it means that everyone kind of got the theory at one point, but the pace of sharing was like how quickly you can get people on a face to face or a phone call, but I'm talking about the old days, like to the production system and on. So you had to trust people to implement first principles in some consistent way, but you really didn't have a full real time feedback loop like the world we live in now. So I think that's, that's pretty awesome to hear. How does that live in a regulated environment that introduces its own set of constraints and important safety and validation aspects to this universe? Daniele: Yeah, I would not say that we have solved it, right? But I believe that I don't think Natan: anyone has solved it completely. Exactly. Daniele: That's why I want to be also, I mean, honest and also humble to say It's something that the industry needs to also solve, most likely also together. That's why Russia is very active in different environments, working very closely with universities, with institutions like ISP, World Economic Forum, and so on, to really create a cohort of people which are rethinking the way we should validate our systems and solution. I think that's the next transformation. On the quality side. On the quality side. Why? Because I think we will build more and more capabilities. Digital analytics capabilities as a service, right? So more and more people will be able to build, use models. And if you are able to build an application in two days in future, this means that you need to be also very fast in validating this application, right? If the validation takes six months, And you're building an application in two days. You can imagine how frustrating this can be for the people. Natan: Absolutely. And you know, one would want to assume that if you're building an application in two days, the reason you're doing it is because you're answering a need that comes either from operations or quality. In fact, like increasing throughput and product quality. And if it's not matched with the ability to validate it in a, of course, safe and up to spec fashion, then we're missing something. Yes. And, you know, you told me a story that was super interesting to me because, you know, we kind of managed almost to go through a full episode without going deep into Gen AI, which I feel is, uh, We're all like super tired from hearing, and I think it's actually a good thing, like a little bit, I know we chatted about this, the anti hype that we're feeling, because the dust is starting to settle, okay, we understand what this GPT stuff can do for us, what the image engine can do for us, we get wacky. Pictures of, you know, squirrels turning into Superman. And okay, that's nice. But what does it actually do for us in operation? So of course, you know, that's not the purpose here. But like on what we just discussed, you had a really great story. You have to use a Gen AI that meets quality and meets like Real impact to the business. Do you mind sharing that a little Daniele: bit? And I mean, you were a bit provocative, right, in saying, I mean, is Gen. AI dead, right? Yeah, I mean, dead in the sense that At least the hype, right? The hype is dead Natan: because, you know, I do this like tests when I'm speaking or, you know, talking this like, who is using Gen. AI or some chat GPT on a daily basis today. At this point, and I don't have an exact chart, but like, you know, it went from 10%, 20%. Now I think it's like over 80 percent people are like admitting that this is already something they do as a common usage pattern and like the way they interact with the data, internet, et cetera, et cetera. The thing I'm not hearing yet is like, now I'm using it in one of my production systems or B2B systems to change how I work. Daniele: Yes, Natan: and I think this is the next frontier and there I don't think it's dead. I just think we're not Daniele: working on it. I agree. My sense is that I think the more basic way of leveraging Gen AI comes now a bit to a plateau and this means basic means for me type of assistance. Functionality. So I'm leveraging Gen AI to summarize email content. I'm leveraging Gen AI to propose some emails, record and summarize meeting methods. a bit more the admin type of services. I think it's spread across all the companies at highest scale. And I think it's good. But I don't think that there is the highest value that we can generate through Gen. AI. That's why I believe that we'll see in future more really, how can we leverage Gen. AI to again, create a different way of working. In my organization at Roche, we are obviously also doing the more admin, basic, Type of services and support is great, but we are also looking at areas, for instance, in our regulatory functions where people need to really work across different systems, needs to write filing reports. It's a quite manual work when we are leveraging GenAI to really augment the people to focus on value adding activities. So we have GenAI, which is really proposing. So, this is really the proposed document that we can send out. Right. So I think it's really completely different way of how we are looking at specific processes. Natan: So this specific example, The previous state would be that a human would sit down with access to various systems, maybe knowledge of a specific event that happened, understanding the SOPs of the company. Now we need to do a report. That's the action. And so that would have been a manual, very Daniele: manual task, very time consuming, right? Reading through many documents, trying to summarize those documents, put different information from different systems. At the level that meets what regulation requires. Absolutely. And now you Natan: can have a Gen AI help make the, call it the draft. Yes. And that substantially cuts down that portion of time. Yes. And it's probably not doing a worse job necessarily because it relieves the, you know, the burden from the human. But what's interesting about this is it keeps the human in the loop. Daniele: Absolutely. Absolutely. I mean, at the end, the human is still required, again, it's a proposal, it's a first draft. Which cuts down significantly the time to deliver the report, but does not replace the person. It's really the augmentation and I think that's the interesting part of GEN AI, right? Where we were debating, I mean, do we still need specific people, specific roles? I still believe that it's required, right? Their human interaction is required, not only from a regulatory, but I think from a value perspective. It's just, how do we augment people to then be more efficient as well and effective in You have to Natan: keep your people. We're missing so many people in operation and the cost of recruiting, training, retaining them is so high. Daniele: And again, there are very complex processes, right? If you're looking at the way we're investigating, let's say, quality deviations, very complex problems, having Gen AI really tapping into all kinds of information. And combining those in a very comprehensive way, it's really helping to also identify the pure root cause and not only a symptom or a cause. Natan: Yeah, we talked a lot about impact today. And the thing that I see a lot is that people ask, what is this value of all this GNI? It's like, at the end of the day, it's very simple. It's just giving people back time. And giving them the time to do other value added thing. And then, you know, in the future of the stream, where is it going? You know what's ironic about it? It's, it's kind of like saying, you know, the future of Gen AI is human. Because if you don't have the human in the loop, then you really can't, at least in operations, you know, I'm sure there will be amazing Gen AI that will like wow us and their ability to create all sorts of kinds of media and content. Change the way we, I don't know, interact with each other. Like all that will happen. Sure. Daniele: Yes. Natan: But in the here and now of like, you know, the work of making drugs or making a car or whatever it is that helps us survive as humans, the future of GNI I predict will be very much requiring a human in the loop. I like a lot the Daniele: slogan, the Natan: future of GNI is human. Daniele, thanks so much for joining Augmented Ops. It's been really fun. Announcer: Thank Natan: you so much. Thank you Announcer: so much. Thank you. Thank you for listening to the Augmented Ops podcast from Tulip Interfaces. We hope you found this week's episode informative and inspiring. You can find the show on LinkedIn and YouTube or at tulip. co slash podcast. If you enjoyed this episode, please leave us a rating or review on iTunes or wherever you listen to your podcasts until next time.