Episode 166: Turning Enterprise AI Into Real Results | Larissa Schneider, Cofounder and COO of Unframe

Welcome to the next in time podcast where we explore the fascinating depths of how people have the potential of impacting the world with a mission and vision of their project. Join us on this audio journey as we uncover the hidden gems of one's vision. Delve in the thought-provoking discussions of why they're pursuing it and see how they're going to make an impact. If you're a curious person, this podcast is your go-to destination. Welcome to the next in time podcast. I'm your host ST and today our guest is Larissa Schneider who is the co-founder and CEO of Unframe where she helps large companies bring AI out of the idea stage and into the real use across their operations. So Larissa, welcome to show. Hey, thanks for having me. I'm excited to be here. Yeah, glad to have you on. I know you've you're based out of Germany and one thing you people in Germany love to do is try to solve complex engineering problems. Exactly. Yeah, I mean, you could combine that. I think we have a pretty global approach to the company and there's definitely a German side to it. I would agree. I mean, it's so German that you just have to basically, all right, here's the problem. All right, let's get it solved and that's why people always have a positive view of Germany in terms of being this very strong engineering powerhouse. Yeah, no, it makes sense. We tried to make it as easy and efficient for our customers as well. And so just out of curiosity. So why do so many AI projects fail and like what actually helps them succeed? Interesting. I think I would say there's two things that I often see which don't really call for a lot of success. One is that customers that are the pretty experimental, I would say, and they want to try a lot of things. They kind of go product first, like what cool LLM is out there? How can I train my own model? Like what type of point solutions or new startups pop up and they just take a very narrow thing that has great technology and then try to find a home for it within the business. And often that fails because they don't really start thinking like pain point first or like real use case first, which is really, really important. And then the other thing is that with AI, we've kind of everything is new, right? Like everything is that we've known from the world, the software world, the SaaS world, everything's been turned on its head over the last couple of years here. And often people don't really know to measure what good looks like. So we always start with our customers like what are you actually trying to achieve? And what are the measurements or the KPIs that you can put in place to make sure that the test you're running, it doesn't matter who that is with that with us, is that with a different company, are you trying to build something yourself? Like trying to figure out what you're trying to achieve and then measure against it. And if you see success, that's perfect. But if not, just like make sure you know what you're tracking towards. Yeah, I get it. It's not easy to be able to find out because the one thing they keep saying is that the AI bubble is about to burst or they say, hey, you only got five years, then someone told, I listened to another episode of a podcast somewhere they said, if you've got five years to adapt or else AI will completely consume your entire system. Is that really true? Is it just just like a myth? I mean, what do we know? We're kind of in the first inning of this whole thing. And so there's so much more to come, I think. But what I can tell you, we're definitely seeing, are kind of those full stack AI companies. So think about new, like let's say in the legal world, a law firm comes out that is not just telling technology, they're building a whole new law firm with AI processes, AI assistants, like everything is AI native there. They will be able to operate significantly faster and cleaner than any of the large incumbents that already exist. So if you are a legacy law firm, you will want to make sure that you are improving your processes as quickly as possible in order to compete with these new companies that are coming in before they take all of your business and you're not able to act fast enough. Like how do you normally balance like day to day, like how do you balance long term vision with the day to day reality of running like a big AI company? Yeah, that's interesting. I would say it's good to think ahead of where you want to go, but don't let that consume you. You know, some of the enterprises we work with, we see that they do like this five year plan in AI right now. That's absolutely insane. We don't know what is ahead of us. And so you want to make sure that you go step by step with quick wins. And so that's what I would say, we do internally as well. Like obviously we had the luxury I call it of starting this company when we already knew about the LLM. So from day one, as we build our frame, this was part of everything that we designed in house, but also the way that we operate with our customers. I see that pretty pretty similarly. Have the quick wins filled on top of it. Do more of what works. Friends and repeat. Don't try to tackle everything one huge huge transformation positive as one because you will be discussing forever and you're not going to get moving. And so before we go into the company on frame AI, tell me more about who you are, where you come from. How did you even end up in the AI space? Yeah, sounds great. So it's you said I grew up in Germany. I was always really, really interested in everything that is fast moving tech. I spent some time in the Bay Area, really learned to entrepreneurship global marketing and so on over there. Join to start up. They were looking to open an office in Europe. So they moved me actually back to my home country and I was more in the cloud infrastructure side of things there. Ultimately during COVID, I decided I wanted something like a different part of tech actually. I joined a cybersecurity company called No Name Security, which was co-founded by Shy, who is now the CEO of Unframe. So we Shy, I, and I all met at No Name Security and Cybersecurity. And it was like that moment, the chat GPT moment happened. You know, I'm gonna change everything in our personal lives and we're still thinking about really old legacy software in our professional life. And we said that there must be a better way of doing this and bringing real value to enterprises with AI in the business context. And so it was a good moment to leave behind the cybersecurity world and move into AI. We co-founded the three of us together and have grown the business from three founders to close to 100 employees now. That's a lot. That's a lot. And just like how many years, how many years ago did you start? But under two years. Under two years. It went from zero to like 100 employees. Which is not very, not very common in many startups these days. Everyone, everyone's going for like, all right, two to ten employees max. Just get it done. Like, that's the cool success stories that you see on LinkedIn. And of course, that can potentially happen. Can that really happen when you are selling to Fortune 500 enterprises with global footprints? I don't think it's realistic, especially if you're working on real business, critical solutions and efficiency. I think you need more people and also have the trust of a large enterprise. Yeah. I mean, do you normally like work with the enterprise? How do you just have the work? You work with the enterprise and you try to find a solution. You just implement the AI solution and whatnot, right? Correct. I would say when we first started, people were like, okay, I need to do AI, but we don't really know how to start. That's gone a lot less over the last like 12 months or so. Now, we usually have fun tank with someone who's part of the CIO org. Like someone in innovation, business transformation, emerging technology is one of those functions. And they usually have like a very, very long list of AI use cases that they're trying to tackle. And they're actively looking for a partner to get that done with. Like, is it true that, you know, a lot of people in Germany, they go to the bay to get all their all this knowledge about technology and then use it to implement it across the world. I don't know if that's specific to Germany. I mean, I'm I've met a lot of German founders over in the bay when I was living there. So yeah, I mean, that's a lot of Germans out there can tell you that. At the company we are, had quartered in California. We have teams all over the US. I am based in Germany. We also have a large R&D center based out of Israel. So we're pretty, pretty global footprints. But I would say most of our customers right now are largely US based. And I hear that your co your co founder your domain co founder CEO of the company is based out of Israel, right? Yeah, correct. I mean, both of us spent, I would say like 50% in Israel and Germany where we have sides as well as the other part in the US without customers. So a lot of travel. So how do you tell me more about him? Like who who he is? How do you even get started with on frame? Yeah. So Shilevi is the CEO, my co founder. And so he has been an entrepreneur multiple times very successfully. It is that previously he started, no name security, which was the leader in API security markets from 0 to 200 people in four years roughly and was acquired by Akhamife 500 million in 24. So it was a pretty good ride. As I said, we had this idea for Unframe while we were kind of in the cyber security world, but we really saw the AI moment happen in front of our eyes. So we decided let's do something that's the shift from one segment of tech to another one. And so yeah, it turned out to be a really good moment to do that. Got it. And so now I'm now fast forward to Unframe AI. So who are the types of co-ph? So are you more like a solutions provider for these Fortune Fiber and companies? Or is it more of a product-based thing? Or is it like a service-based thing? Now definitely more of a product. So we obviously given the nature of our business and the state of AI right now, we are very consultative and we have a lot of services component to it, but it's not something that when our services company is no SOW kind of engagement with our customers, they don't pay for it. But our product is a platform. It's a SaaS subscription. So the way that you know it from based likes and you decide if you like it, you continue renewing for multiple years. Or even some stage, something else happens and you want to orient yourself differently than there's no harm. It's a subscription. Makes sense. That's interesting. So what are the types of solutions you normally provide? What types of industries, what types of solutions and what types of Fortune Fiber companies have you worked with? Yeah, sounds good. So we're pretty industry agnostic. But we see some industries that are just leaning in and moving significantly faster than others. So where we're very successful financial services, insurance, real estate, retail, manufacturing, those are super interested for multiple reasons, either because of high competition, high pressure to get the AI strategy rolled out. But also for some more technical reasons, like they have a lot of legacy systems, desperate systems that the employees work with. Think about someone on the admin side who has like six tabs open to get one simple job done. Something like that is a prime example for having AI help you reduce that manual labor. Or you have a lot of unstructured data. So when that happens, there's a lot more that you can do with AI right now to help in those processes. Usually our customers also have a big international footprint. So think about they have data in lots of different languages, lots of different currencies, metric systems, and so on. And so their AI is really super, super helpful to bridge the gap across those different global teams. And so I see that you're in areas like data engineering, finance, IT, legal operations, and sales. So we usually break down the starting points like the solutions that are a good fit for our platform to three categories. The first one would be observability and reporting. So think about all of the BI analytics and search type use cases and that can be from across all the different departments and verticals that you just mentioned. The other one is extraction, abstraction of unstructured data. So any type of use cases that would be your financial statements, Excel files, reports, contracts, and so on. And then what we call automation agents. So all of the proactive AI workflows that can help with decision making and actions. And so yeah, those can apply to many different industries, but those are usually really, really good starting points and it's a matter of prioritizing them for the enterprise and finding a good way to see how I can bring value to yourself. And it looks like you also have, it's also powered by Galileo and synergy. So those are two of our solutions that we have named because they're very repetitive. So those are our own terminologies. One of them is on the enterprise search and data correlation side of things. And the other one is an IT operation solution, which is a use case that we see very often with data from across different systems like ServiceNow, Gerard Confluence, Dinertrades. And so on most companies use these systems and so are sometimes struggling to have them talk to each other. And so how do you, let's just say for example, I am, let's say we are a Fortune Fibonid Company who specializes in manufacturing, right? Or let's say we're trying to sell consumer goods. And so how does your solution be a helping hand to that, especially in areas that are like less pro, like techno IT and software tech oriented? Like how do you provide that solution? How do you try to fix that problem? Yeah, so usually we would talk either with the CIO. Let's, for this case, let's imagine we're talking to the CIO of a manufacturing company as you said. And we would try to figure out what is their main pain points right now. And they might say we have a lot of workers in our factory for struggling with the instructions booklets that they're getting. And so they want to get learnings out of it. They want to be able to search in natural language and get instructions very quickly. We would be able to help them set up a solution for that on top of our platform with the building blocks that we have. And a matter of let's say a week, usually we either get an integration into an existing system that we already have or maybe some sample data, which would also be really, really good. And if we have those things, we'll put together the solution. It will be complete production ready for them to try within a week. And they can see if they like it. If it brings them value, after the POC will move to licensing, which are per solution per year. And over time they might say my finance team saw how you work with our factory workers. And they also have a use case or our sales team has some kind of AI use case that could also be interesting. And we're very flexible in adding those additional use cases to our platform. And that's a really, really interesting part for us when customers that see value and one division grow with us and we help them across the business. Makes sense. And so what? How do you find those? So how many clients do you have right now in total? If you had to put it all together, all the companies worldwide, how many have you found? How are you able to maybe find that approach? Yeah. So we started, as I said a couple of years ago, still in the double digits. Many of our customers are using multiple solutions of ours, which is very exciting. It's a huge expense for us. It's a mixture. Either we work with our prospects when they're not customers yet and do a lot of workshops and brainstorming sessions around what a good starting point with AI would be. Or they already have a very clear idea and they have a use case prioritization list out of which we start working through the list. Both of those happen. Okay. And so how many, how many total numbers in terms of when you started this company to what to now, like how many clients have you, have you service so far? Yeah. I think that double digits, we don't publish exact specifics right now. But within those customers, obviously they're large enterprises so there can be many hundreds and thousands of users in each of the customers and multiple solutions. I think you're trying to make your solution very unique because every company out there, they have their own AI teams or they try to have their own like the internal teams getting their, getting it done. How are you trying to be the person trying to be the, how are you trying to be the person providing that solution that will make it, you know, make that make it, make it, make a difference. Yeah, absolutely. Of course many companies have AI teams and that is great, right? The AI teams, we would say it makes most sense if you put your own person on the own IP that you're trying to create, right? That is the type of use case or the type of development work that they should be doing that moves your company forward. And often that's something that's custom-facing or generating your own IP. But there are usually many, many other use cases that don't really require internal teams. Like let's say an example, in most cases, the company still buys an ERP or a CRM like Salesforce without even considering buying that, you know? And in the same way, if you are, let's say, a car manufacturer on automotive company, you would want your own internal teams probably focusing on creating the autonomous vehicle rather than creating an HR ticketing software. But there could be a lot of efficiencies gained through having AI processes and agents and workflows in the HR department and that's a great point to work with on-frame. Yeah. And so, where do you see this going in the next couple of years in terms of how the types of clients are targeting? I think we are going to continue targeting similar to now, like global enterprises, Fortune 500, global 2000. Those are great companies for us to partner with. And what makes me really happy is when Tesla has come back for use case number two and three and five and ten that shows that they are getting good value from working with us and keep expanding into different areas as well. And so, how are you doing this differently? Because, you know, there are so many people out there providing AI solutions. How are you trying to be the standout? Yeah. I mean, there are a lot of companies that are out there that are providing point solutions, right? like you have one software for one use case. But if we're thinking global enterprise, I talked to a Wall Street bank not long ago and they told me to have a backup of like 1,600 something, AI use cases right now. She told me she's not able to actually find a point solution for every single one of them. And she also is not able to have all of these built-in house. So she needs a scalable alternative. And the platform like ours that can be so tailored towards the specific use cases that an enterprise has and do that in a scalable way is actually a really good thing for her to get through. And so how has your international background shaped the way you lead an operate? Yeah, that's a company overall. From day one, we were very, very international. We started even as the founding team, as a set, we're all in different places. And so that is how to get the foundation of working with large enterprises that have global footprints in our DNA from day one. When you work with enterprises, you need to consider different time zones, different languages, different type of legal setups, super important. And as having now, I mean, obviously, the US is our biggest market. But also European, UK, Middle Eastern sides expanding into Asia as well, that really helps us to understand how the global enterprise operates. And so how did you and your co-founders just had a team up and what strength does each of them bring to the table? So both of them actually have a very, very technical background, which I absolutely do not. My background is more on the go-to-market side. Obviously, Shai has found his companies before. So this was the natural next step for him being a great business leader. Adi is extremely technical. And also has that mindset of scalability, which is really important because we build a business that, from day one, was supporting not one product, but dozens of products. And so that needs a very specific setup as you're building out an engineering and products and to support function. And that was something that the two of them really brought to the table. I focused on the go-to-market side early on, founding partnerships as well as getting us ready to go to markets and do our initial launch out in the world. And so yeah, that's how we got together. We're extremely different all three of us. But that's, I think, important when you try to find your co-founders. And so where did you meet again? No, I'm secured. No, I'm security. No, I'm security. Got it. Interesting. Yeah. Yes. And so what-- how do you-- so do you see this-- do you ever see yourself getting into the big company-- I almost say Fortune 500, but let's just say the top 10 companies on, for example, Tesla. Would you see yourself trying to provide a solution in that realm? Yeah, it's possible. Absolutely. I don't see a reason why we wouldn't. We would probably not shoot for the in-vehicle type of-- I mean, and also Nvidia, because I think one of the things you guys are-- one of the things about Nvidia is that there's just so many. There's a lot of things. I mean, it's already becoming like a very valuable company right now. It's on its way. And it's like, they also have their own internal AI teams. I'm going to-- I'm posing a question right here. So how would you approach that? I mean, actually, I don't think we have a single platform that does not have the owning IT. And that's totally fair. I mean, I said that before, you won't have an enterprise that only does build or only does buy. It is a very logical thing for them to do both sides. For the right use cases, I would say 20% probably should be built for your own IP. 80% is probably the software requirements that are not internal IP in many companies, especially traditional industries. And you can work with a lot of external partners at point solutions platforms, service providers, to get those use cases done. And so there's a lot of internal efficiencies that we can be helped with AI. And any company, even the very, very big ones, have internal inefficiencies that they often try to figure out with software. All right, cool, Larissa. I think we're going to wrap this up by saying, thank you so much for coming on the next-in-time podcast and wishing the best of luck with on frame. Thank you so much for having me. Thank you for joining us on this episode of the Next-In-Time Podcast. We hope you enjoyed diving into the intriguing vision of our guest today. Don't forget to subscribe and follow us on social media to stay updated on future episodes. If you have any suggestions or feedback, we'd love to hear from you. Until next time, stay curious and keep exploring.