theCUBE + NYSE Wired: Actai Maui | Helena Zhang, Osly.ai

Clips
More from theCUBE + NYSE Wired: Actai Maui

Helena Zhang

Founder

Osly

play_circle_outline Startup showcasing natural language to code

play_circle_outline Team dynamics and motivation for project

play_circle_outline Race to AGI and overcoming blockers

play_circle_outline Exploring the Crucial Role of Systems Thinking and Interdisciplinary Knowledge in AI Innovation

Info

Helena Zhang, Osly.ai

Helena Zhang

Founder Osly

search

>> Hello everyone, welcome to the CubePod. I'm John Furrier, host of theCUBE. We are here in Maui for the Acti Global event. Behind us is the beautiful ocean where everyone's kitesurfing all day. Talks in the morning, when the wind's down. A lot of entrepreneurs, investors, leaders from all walks of life, athletes, innovators, artists, conservationists, and of course we got the tech folks here, the technology folks from, again, all walks of life. Lou Zang is here, she's an entrepreneur, recent Stanford grad, exciting startups. All the buzz is about you here at the event. Thanks for coming on the podcast.

Helena Zhang

>> Thank you.

>> So everyone's talking about you and they're saying, wow, she's going to be the next big thing. And you got great demos because I saw you cranking the demo out on coding while you were getting ready to present your new endeavor. How's it feel?

Helena Zhang

>> It feels great. I'm so excited about the Endeavor

>> And you're kitesurfing.

Helena Zhang

>> Yes.

>> How you feel about the kitesurfing?

Helena Zhang

>> Oh, I love it. I think I'm going to get really hooked.

>> It's an endorphin rush. Big time for me. I enjoy learning, so I'm excited. Where I'm at, I feeling pretty good, but it's like learning anything, right? It's like, okay, just got to get the foundation. As Brandon who's just on talked about, he's a pro and the old expression and tech pro tip. We have pros here at the event mingling with us. Tons of pro-tips. What's your favorite pro-tip that you've heard?

Helena Zhang

>> Yeah, I got to say there's been a couple. One athlete was talking about making risky decisions with no emotions involved and I love that. And then the canonical lesson of how slow is fast I really love. Super relevant to what we're doing. So we're excited to go to market, but we need to go slower to make the tech perfect. Those are probably my top two. Yeah.

>> Talk about what you're working on because you have a demo. It's in beta as a waitlist. Talk to about your opportunity.

Helena Zhang

>> Yeah. Yeah, it's in beta right now. We're super excited. It's a natural language to code to workflow platform. And the goal behind this is to empower everyone who is non-technical to leverage AI code tools to automate repetitive tasks in their life, just because time is probably our most valuable asset and people spend a lot of time on very repetitive tasks for work. So it'll be super empowering if we can get people just creating these AI code tools to streamline workflows they have to do every day and save them time and let them just be creative.

>> You and I were talking yesterday about this, and I want to get a little bit dig deeper into it. You said you're super excited about your team.

Helena Zhang

>> Yeah.

>> Talk about the kind of people you're working with right now. What are they motivated for? What's exciting you about the ensemble, the people around the project?

Helena Zhang

>> For one, we're super eclectic. I had no idea how we all came together. It was all so serendipitous. And the other big point is everyone's really kind-hearted. At the same time, risk-taking. And that's why I love this group so much because there's this charge into unknowns, but also everyone hypes each other up like crazy. And I really admire and respect that. To go a little bit into the team, my co-founder, Jacoby Haskell, he is just wrapping up Brown. He's actually graduating this weekend, so I'm super happy about that. And he's just the best vibes ever and super brilliant. And he's been on EST, I'm on PST and it would be 12:00 A.M. and 3:00 A.M for him. And we would still be working and I really admire the grit. We also have Yossi who's a sixteen-year-old and he's a special kid. He's from Saratoga, Florida, and I've never met anyone as ambitious as him because he literally thinks they're going to create AGI. And then we have an older person who helped Grok speak at XAI, Benjamin, and he's a very critical logical thinker. And we had these other folks I know from school on the team. And yeah, they're just so-

>> Lets go.

Helena Zhang

>> Yeah, yeah,

>> They're ready to go.

Helena Zhang

>> And they're so passionate about the mission, which I think is super important because I'm so excited to make this a reality for the 35-year-old who's never taken a CS class but has repetitive tasks and wants to automate at their corporate job.

>> I was talking to someone coming out of college, a mechanical engineer, grade school, and he was trying to find a job. He was aerospace engineer, super brilliant. He could code, he could do everything. And it was hard for him to find something that he really liked. And then he got into drones and he wrote the kernel for the drone to go longer distance. And he's not a CS major, he's just the problem solver. But he found that the tooling they were using just allowed him to do it. And I said, that's awesome. And he goes, yeah, we're the Lego blocks generation. And I'm like, okay, cool. I've been using that. So the composability now and with the tooling, and it also came up, Brett mentioned something yesterday I thought was compelling, where a lot of the cool stuff coming out of the groups that are cranking on AI is like the builders. There's all these tooling, AI tools. It makes tooling easier. So literally building a tool out of thin air, so to speak, I need a tool to do this and I got to make it work for this. You can actually create custom tooling. This is like killer innovation. What's your reaction to that? And is it just oxygen to you guys? Because when I was in school, they didn't have that kind of stuff, so you had to actually build it.

Helena Zhang

>> Yeah, no, it's something we definitely don't take for granted. There's a lot out there, and the best way to do this is to leverage as many as possible. And it's kind of like when you talk to other people and you try to learn from them, and it's just a coalescing of different things. Dario from machines, Grace talks about this. AGI would be the coalescing of different models and infrastructure, and I think it's just creating a piece of the pie there would be super valuable.

>> So race to AGI, are we going to be there soon?

Helena Zhang

>> Yeah, I believe in it. Yeah.

>> What's the blocker in your mind?

Helena Zhang

>> I think the blocker is making the model smarter and that requires data and infrastructure. And for instance, right now, my favorite book is Anna Karenina. And I don't think AI can write a novel like Tolstoy and have it be so creative, but I think we'll get to the point because we've seen AI be self-learning and capable of improving itself. So I think we'll get there soon. Yeah.

>> What are some of your goals right now? You obviously have a small team, you guys are putting the foundation together. What's it like? What are your goals? What are you working on?

Helena Zhang

>> Yeah, the big goal is to make a great product. And personally, I just want to keep the team inspired, bring on new folks on the team. We want to grow as fast as possible. There's some crazy growth stories in the valley of companies that shoot up within a year or two, and we just want to get the strategy right. And there's actually, I met so many people here who are so helpful. Like Michael, who's a marketing pro, I'm so excited to connect to Mark from Intel and see what we can do there.

>> He offered to be a customer out of the gate.

Helena Zhang

>> Yeah.

>> Yeah. We'll follow up.

Helena Zhang

>> Yeah, we'll follow up.

>> Get that . All right, so you got to go to market fast. Cool. Startups are faster now. How are you managing that? Because I love that line. You got to go slow to go fast and kite surfing. I had the death grip. Brandon was talking about the death grip. I had a little bit of a death grip. I solved that problem. But there's a danger to go fast, but you want to go fast. What's your mindset like? What's the vibe of making it go faster without going too fast where you kind of crack? Are you waiting for the straight and narrow to go straight? Are you guys navigating the product? Are you looking for certain use cases? What are some of the things that's going to help you go faster? What are you going slow for? I mean, it's like driving a car. You don't want to go around the corner and go off the cliff. Is it use cases? You looking for more data? What are some of the things that's going to help you go from slow to fast?

Helena Zhang

>> So we're operating a general purpose engine. We can really automate any workflow and we want to leave it open-ended at the beginning to pinpoint what is the category of workflows people are most excited about creating. And once we find that we're just going to go fast on that. But right now we're still in the slow process, kind of waiting for the wind and just seeing what customers want and love. And then I think we're going to drill that. Another big thing we do is we definitely leverage AI to write code for us, but one thing we do is look at every line that the AI writes to make sure that the system runs and it doesn't break anything because AI is definitely known to write code that doesn't make all too much sense.

>> Yeah, you got to code reviews.

Helena Zhang

>> Yeah. So we do our code reviews and we're very meticulous and careful about that.

>> Yeah, I want to get your thoughts on that because, first of all, the tooling is great. Coding assistant now goes to coding, then you got to have the code reviews. There'll be code review agents. Then they got to have look at the agents. So you need an agent for the agent to do the code reviews. And that's impacting kind of the creativity and the role of the computer scientist. Now, Sergey Brin was quoted last week saying that it's best time to be a computer scientist, which I agree. By the way, I think it's a great time because there's so much opportunity. So the question is they just graduated from Stanford. My son just graduated last year, CS major, and I was saying to him, are you worried? That's what I asked my son. I said to him, are you worried that the CS degree might be outdated. And he's like, "Yeah, because the stuff I'm doing now has nothing to do with school, blah, blah." I did learn problem solving. But coming out of Stanford, is there stuff that you look back and saying, hey, the CS programs are... What they teach is changing.

Helena Zhang

>> Yeah.

>> What's your view on that? Because it is dynamic.

Helena Zhang

>> Yeah.

>> Should they go more theory and coding is going to be good-

Helena Zhang

>> So I think the systems classes are so beneficial. We've actually written a paper on this, humans and AI collaborating, what that looks like and swaying up what humans are good at and what AI is good at. And humans are the creatives here. And humans are the ones that define logic and has niche knowledge that AI currently doesn't have. For instance, you can get AI to create a game, but people who are really good at the game say poker, they have inside knowledge that an instruction manual won't have. So I think that's where the humans stand. And then the AI is great at reading docs, writing good syntax, and you just need the human to guide it a little bit right now. But yeah, I think it's super exciting because these roles are blending.

>> So systems is important, systems knowledge. So systems thinking is a big deal.

Helena Zhang

>> And first principles like the why to dig a layer deeper and to think about architecture I think is super important. And I also think there's a huge emphasis on just studying very interdisciplinary things now in addition to CS, because I've painted all my life actually, and for me it's always been pulling ideas from different places in my life. And I had this art history professor at Stanford, his name's Alexander Nimrov, and he spoke so beautifully about art and the world because he would connect these really far away things and he would create a story out of it. A good example is he talked about Bentley, this photographer photographing snowflakes during the Civil War. And people were like, "Bentley, why are you doing this? There's people out there fighting and you're taking photographs of snowflakes." But then he took this ear book of snowflakes that's currently at Cal Academy of Sciences, I believe, and then connected it to a photograph of the fallen soldiers. And it's like every snowflake is unique and falls and disappears, and he connect the two photographs together.

>> That's hard.

Helena Zhang

>> And I think just creating these faraway constellations is super important because so many tools out there. So your job is to find the right tools, connect them together and be creative about it. And yeah, I think art personally has helped me a lot. I also spent a long time studying neuroscience, so that inspires me a lot. Just how the brain works and how we can put that into our architecture.

>> Neural networks are hot right now.

Helena Zhang

>> Yes, yes. Neural networks are very hot right now. And it's actually through neuroscience that I got into ai. I got involved in this project that was generating synthetic pet images from MRI images since radiotracers are inaccessible in rural areas. And I just thought it was so impactful. And yeah, there's just lot-

>> Isn't it funny how the universe gets you somewhere because what draws you in might not... It's not the path that everyone thinks is the path. Go to get a CS degree, code, build a product, get an MVP. The old models are evolving to just solve a problem and get it out there and bring everything you can bear to the effort.

Helena Zhang

>> And I really think at this point, if you're determined to do it, you can do it. It doesn't have to be something you've known since you were five years old. You can learn it. There's enough knowledge out there, especially with AI.

>> With AI, everyone can level up. You don't have to have pedigree. And although Stanford's got great pedigree, but that helps. But I mean, anyone can get into computer science if they think like one, there's no really a degree needed and the aperture is opened up.

Helena Zhang

>> Yeah, yeah. Well, we think about systems a lot. Our name Osly, O-S-L-Y, is actually a play on operating system. That's what OS stands for. And we're thinking about the compiler as the LM and prompts right now.

>> Yeah, a compiler native to the LLM's or the front end where the business logic is. Sounds like something you're going after. Business logic is where the action is.

Helena Zhang

>> Yes.

>> So talk about your project, people watching, if you're watching. This is an area that people do talk a lot about. We talk about a lot in our podcasts about the undifferentiated heavy lifting. Andy Jassy's favorite line. There's no compression algorithm for experience. He mentions that too. Toil, grunt work. These repetitive tests are productivity killers, opens the freedom for the user, customer. You can be more creative, bring the human into the problem. What use cases are you looking for while you're in waitlist mode for you're bringing in selective customers, what are you looking for? What's the ideal use cases that in case anyone's watching wants to get involved? What's the desire for, give me your use cases. What's the list look like? What's the criteria? What's the focus?

Helena Zhang

>> We really love doing enrichment workflows. That could be lead enrichment, deal enrichment, finding information out there and pulling it all together. And some of it looks like connecting different data sources like CRMs, populating the CRMs. It could look like going to a browser, doing some action on the browser, like scrolling to the bottom, finding the terms of services, summarizing that, and then doing some action on top of that. So for us, it's just connecting different tools and utilities together and taking actions on top of that, which is a workflow that people do. We're making the bet that we can do this straight with natural language just because people love chat interfaces so much. There's some other approaches here. You can do a screen recording and then use screenshots to automate the workflow. But we think that with natural language, you can compose building blocks and then string the building blocks together and end up orchestrating even more complex workflows with more reliability as well. And then, yeah, reliability.

>> So knock down some of those. But then make it smart enough to be generative. So generate workflows where there might be new integration opportunities.

Helena Zhang

>> Yes. And our big vision for this is we want to be the enterprise workflow engine for big enterprises because right now, they're not so agentic. I want to-

>> They want to be.

Helena Zhang

>> Yeah, there's a couple in my mind right now, but they're like a chatbot and rag on top of it, and we just think that there's so much more that can be done there.

>> That's kind of the low hanging fruit. It's like kitesurfing beginner class.

Helena Zhang

>> Yeah. Right. And we're just so excited to see if we can take it one step further. Yeah.

>> Well, I'm super excited for you, and you're such an inspiration here at the show. I'm really impressed by you. And again, I'm so happy for you and we got some use cases. We'll try to bring them in. And good luck and be certainly talking soon. Again, thanks for coming on.

Helena Zhang

>> Thank you.

>> Appreciate it. Thank you. All right. All right, we're here talking to all the entrepreneurs. We've got investors coming in. Also, senior members in the community are coming together too. So it really is a mixed community. Young, old coming together. Systems people here, they think about consequences, first principles, second order consequences. It's a systems thinking revolution that's happening with AI. And again, we're here in Maui with a beautiful background, sun setting. Thanks for watching.

Helena Zhang, Osly.ai

search

>> Hello everyone, welcome to the CubePod. I'm John Furrier, host of theCUBE. We are here in Maui for the Acti Global event. Behind us is the beautiful ocean where everyone's kitesurfing all day. Talks in the morning, when the wind's down. A lot of entrepreneurs, investors, leaders from all walks of life, athletes, innovators, artists, conservationists, and of course we got the tech folks here, the technology folks from, again, all walks of life. Lou Zang is here, she's an entrepreneur, recent Stanford grad, exciting startups. All the buzz is about you here at the event. Thanks for coming on the podcast.

Helena Zhang

>> Thank you.

>> So everyone's talking about you and they're saying, wow, she's going to be the next big thing. And you got great demos because I saw you cranking the demo out on coding while you were getting ready to present your new endeavor. How's it feel?

Helena Zhang

>> It feels great. I'm so excited about the Endeavor

>> And you're kitesurfing.

Helena Zhang

>> Yes.

>> How you feel about the kitesurfing?

Helena Zhang

>> Oh, I love it. I think I'm going to get really hooked.

>> It's an endorphin rush. Big time for me. I enjoy learning, so I'm excited. Where I'm at, I feeling pretty good, but it's like learning anything, right? It's like, okay, just got to get the foundation. As Brandon who's just on talked about, he's a pro and the old expression and tech pro tip. We have pros here at the event mingling with us. Tons of pro-tips. What's your favorite pro-tip that you've heard?

Helena Zhang

>> Yeah, I got to say there's been a couple. One athlete was talking about making risky decisions with no emotions involved and I love that. And then the canonical lesson of how slow is fast I really love. Super relevant to what we're doing. So we're excited to go to market, but we need to go slower to make the tech perfect. Those are probably my top two. Yeah.

>> Talk about what you're working on because you have a demo. It's in beta as a waitlist. Talk to about your opportunity.

Helena Zhang

>> Yeah. Yeah, it's in beta right now. We're super excited. It's a natural language to code to workflow platform. And the goal behind this is to empower everyone who is non-technical to leverage AI code tools to automate repetitive tasks in their life, just because time is probably our most valuable asset and people spend a lot of time on very repetitive tasks for work. So it'll be super empowering if we can get people just creating these AI code tools to streamline workflows they have to do every day and save them time and let them just be creative.

>> You and I were talking yesterday about this, and I want to get a little bit dig deeper into it. You said you're super excited about your team.

Helena Zhang

>> Yeah.

>> Talk about the kind of people you're working with right now. What are they motivated for? What's exciting you about the ensemble, the people around the project?

Helena Zhang

>> For one, we're super eclectic. I had no idea how we all came together. It was all so serendipitous. And the other big point is everyone's really kind-hearted. At the same time, risk-taking. And that's why I love this group so much because there's this charge into unknowns, but also everyone hypes each other up like crazy. And I really admire and respect that. To go a little bit into the team, my co-founder, Jacoby Haskell, he is just wrapping up Brown. He's actually graduating this weekend, so I'm super happy about that. And he's just the best vibes ever and super brilliant. And he's been on EST, I'm on PST and it would be 12:00 A.M. and 3:00 A.M for him. And we would still be working and I really admire the grit. We also have Yossi who's a sixteen-year-old and he's a special kid. He's from Saratoga, Florida, and I've never met anyone as ambitious as him because he literally thinks they're going to create AGI. And then we have an older person who helped Grok speak at XAI, Benjamin, and he's a very critical logical thinker. And we had these other folks I know from school on the team. And yeah, they're just so-

>> Lets go.

Helena Zhang

>> Yeah, yeah,

>> They're ready to go.

Helena Zhang

>> And they're so passionate about the mission, which I think is super important because I'm so excited to make this a reality for the 35-year-old who's never taken a CS class but has repetitive tasks and wants to automate at their corporate job.

>> I was talking to someone coming out of college, a mechanical engineer, grade school, and he was trying to find a job. He was aerospace engineer, super brilliant. He could code, he could do everything. And it was hard for him to find something that he really liked. And then he got into drones and he wrote the kernel for the drone to go longer distance. And he's not a CS major, he's just the problem solver. But he found that the tooling they were using just allowed him to do it. And I said, that's awesome. And he goes, yeah, we're the Lego blocks generation. And I'm like, okay, cool. I've been using that. So the composability now and with the tooling, and it also came up, Brett mentioned something yesterday I thought was compelling, where a lot of the cool stuff coming out of the groups that are cranking on AI is like the builders. There's all these tooling, AI tools. It makes tooling easier. So literally building a tool out of thin air, so to speak, I need a tool to do this and I got to make it work for this. You can actually create custom tooling. This is like killer innovation. What's your reaction to that? And is it just oxygen to you guys? Because when I was in school, they didn't have that kind of stuff, so you had to actually build it.

Helena Zhang

>> Yeah, no, it's something we definitely don't take for granted. There's a lot out there, and the best way to do this is to leverage as many as possible. And it's kind of like when you talk to other people and you try to learn from them, and it's just a coalescing of different things. Dario from machines, Grace talks about this. AGI would be the coalescing of different models and infrastructure, and I think it's just creating a piece of the pie there would be super valuable.

>> So race to AGI, are we going to be there soon?

Helena Zhang

>> Yeah, I believe in it. Yeah.

>> What's the blocker in your mind?

Helena Zhang

>> I think the blocker is making the model smarter and that requires data and infrastructure. And for instance, right now, my favorite book is Anna Karenina. And I don't think AI can write a novel like Tolstoy and have it be so creative, but I think we'll get to the point because we've seen AI be self-learning and capable of improving itself. So I think we'll get there soon. Yeah.

>> What are some of your goals right now? You obviously have a small team, you guys are putting the foundation together. What's it like? What are your goals? What are you working on?

Helena Zhang

>> Yeah, the big goal is to make a great product. And personally, I just want to keep the team inspired, bring on new folks on the team. We want to grow as fast as possible. There's some crazy growth stories in the valley of companies that shoot up within a year or two, and we just want to get the strategy right. And there's actually, I met so many people here who are so helpful. Like Michael, who's a marketing pro, I'm so excited to connect to Mark from Intel and see what we can do there.

>> He offered to be a customer out of the gate.

Helena Zhang

>> Yeah.

>> Yeah. We'll follow up.

Helena Zhang

>> Yeah, we'll follow up.

>> Get that . All right, so you got to go to market fast. Cool. Startups are faster now. How are you managing that? Because I love that line. You got to go slow to go fast and kite surfing. I had the death grip. Brandon was talking about the death grip. I had a little bit of a death grip. I solved that problem. But there's a danger to go fast, but you want to go fast. What's your mindset like? What's the vibe of making it go faster without going too fast where you kind of crack? Are you waiting for the straight and narrow to go straight? Are you guys navigating the product? Are you looking for certain use cases? What are some of the things that's going to help you go faster? What are you going slow for? I mean, it's like driving a car. You don't want to go around the corner and go off the cliff. Is it use cases? You looking for more data? What are some of the things that's going to help you go from slow to fast?

Helena Zhang

>> So we're operating a general purpose engine. We can really automate any workflow and we want to leave it open-ended at the beginning to pinpoint what is the category of workflows people are most excited about creating. And once we find that we're just going to go fast on that. But right now we're still in the slow process, kind of waiting for the wind and just seeing what customers want and love. And then I think we're going to drill that. Another big thing we do is we definitely leverage AI to write code for us, but one thing we do is look at every line that the AI writes to make sure that the system runs and it doesn't break anything because AI is definitely known to write code that doesn't make all too much sense.

>> Yeah, you got to code reviews.

Helena Zhang

>> Yeah. So we do our code reviews and we're very meticulous and careful about that.

>> Yeah, I want to get your thoughts on that because, first of all, the tooling is great. Coding assistant now goes to coding, then you got to have the code reviews. There'll be code review agents. Then they got to have look at the agents. So you need an agent for the agent to do the code reviews. And that's impacting kind of the creativity and the role of the computer scientist. Now, Sergey Brin was quoted last week saying that it's best time to be a computer scientist, which I agree. By the way, I think it's a great time because there's so much opportunity. So the question is they just graduated from Stanford. My son just graduated last year, CS major, and I was saying to him, are you worried? That's what I asked my son. I said to him, are you worried that the CS degree might be outdated. And he's like, "Yeah, because the stuff I'm doing now has nothing to do with school, blah, blah." I did learn problem solving. But coming out of Stanford, is there stuff that you look back and saying, hey, the CS programs are... What they teach is changing.

Helena Zhang

>> Yeah.

>> What's your view on that? Because it is dynamic.

Helena Zhang

>> Yeah.

>> Should they go more theory and coding is going to be good-

Helena Zhang

>> So I think the systems classes are so beneficial. We've actually written a paper on this, humans and AI collaborating, what that looks like and swaying up what humans are good at and what AI is good at. And humans are the creatives here. And humans are the ones that define logic and has niche knowledge that AI currently doesn't have. For instance, you can get AI to create a game, but people who are really good at the game say poker, they have inside knowledge that an instruction manual won't have. So I think that's where the humans stand. And then the AI is great at reading docs, writing good syntax, and you just need the human to guide it a little bit right now. But yeah, I think it's super exciting because these roles are blending.

>> So systems is important, systems knowledge. So systems thinking is a big deal.

Helena Zhang

>> And first principles like the why to dig a layer deeper and to think about architecture I think is super important. And I also think there's a huge emphasis on just studying very interdisciplinary things now in addition to CS, because I've painted all my life actually, and for me it's always been pulling ideas from different places in my life. And I had this art history professor at Stanford, his name's Alexander Nimrov, and he spoke so beautifully about art and the world because he would connect these really far away things and he would create a story out of it. A good example is he talked about Bentley, this photographer photographing snowflakes during the Civil War. And people were like, "Bentley, why are you doing this? There's people out there fighting and you're taking photographs of snowflakes." But then he took this ear book of snowflakes that's currently at Cal Academy of Sciences, I believe, and then connected it to a photograph of the fallen soldiers. And it's like every snowflake is unique and falls and disappears, and he connect the two photographs together.

>> That's hard.

Helena Zhang

>> And I think just creating these faraway constellations is super important because so many tools out there. So your job is to find the right tools, connect them together and be creative about it. And yeah, I think art personally has helped me a lot. I also spent a long time studying neuroscience, so that inspires me a lot. Just how the brain works and how we can put that into our architecture.

>> Neural networks are hot right now.

Helena Zhang

>> Yes, yes. Neural networks are very hot right now. And it's actually through neuroscience that I got into ai. I got involved in this project that was generating synthetic pet images from MRI images since radiotracers are inaccessible in rural areas. And I just thought it was so impactful. And yeah, there's just lot-

>> Isn't it funny how the universe gets you somewhere because what draws you in might not... It's not the path that everyone thinks is the path. Go to get a CS degree, code, build a product, get an MVP. The old models are evolving to just solve a problem and get it out there and bring everything you can bear to the effort.

Helena Zhang

>> And I really think at this point, if you're determined to do it, you can do it. It doesn't have to be something you've known since you were five years old. You can learn it. There's enough knowledge out there, especially with AI.

>> With AI, everyone can level up. You don't have to have pedigree. And although Stanford's got great pedigree, but that helps. But I mean, anyone can get into computer science if they think like one, there's no really a degree needed and the aperture is opened up.

Helena Zhang

>> Yeah, yeah. Well, we think about systems a lot. Our name Osly, O-S-L-Y, is actually a play on operating system. That's what OS stands for. And we're thinking about the compiler as the LM and prompts right now.

>> Yeah, a compiler native to the LLM's or the front end where the business logic is. Sounds like something you're going after. Business logic is where the action is.

Helena Zhang

>> Yes.

>> So talk about your project, people watching, if you're watching. This is an area that people do talk a lot about. We talk about a lot in our podcasts about the undifferentiated heavy lifting. Andy Jassy's favorite line. There's no compression algorithm for experience. He mentions that too. Toil, grunt work. These repetitive tests are productivity killers, opens the freedom for the user, customer. You can be more creative, bring the human into the problem. What use cases are you looking for while you're in waitlist mode for you're bringing in selective customers, what are you looking for? What's the ideal use cases that in case anyone's watching wants to get involved? What's the desire for, give me your use cases. What's the list look like? What's the criteria? What's the focus?

Helena Zhang

>> We really love doing enrichment workflows. That could be lead enrichment, deal enrichment, finding information out there and pulling it all together. And some of it looks like connecting different data sources like CRMs, populating the CRMs. It could look like going to a browser, doing some action on the browser, like scrolling to the bottom, finding the terms of services, summarizing that, and then doing some action on top of that. So for us, it's just connecting different tools and utilities together and taking actions on top of that, which is a workflow that people do. We're making the bet that we can do this straight with natural language just because people love chat interfaces so much. There's some other approaches here. You can do a screen recording and then use screenshots to automate the workflow. But we think that with natural language, you can compose building blocks and then string the building blocks together and end up orchestrating even more complex workflows with more reliability as well. And then, yeah, reliability.

>> So knock down some of those. But then make it smart enough to be generative. So generate workflows where there might be new integration opportunities.

Helena Zhang

>> Yes. And our big vision for this is we want to be the enterprise workflow engine for big enterprises because right now, they're not so agentic. I want to-

>> They want to be.

Helena Zhang

>> Yeah, there's a couple in my mind right now, but they're like a chatbot and rag on top of it, and we just think that there's so much more that can be done there.

>> That's kind of the low hanging fruit. It's like kitesurfing beginner class.

Helena Zhang

>> Yeah. Right. And we're just so excited to see if we can take it one step further. Yeah.

>> Well, I'm super excited for you, and you're such an inspiration here at the show. I'm really impressed by you. And again, I'm so happy for you and we got some use cases. We'll try to bring them in. And good luck and be certainly talking soon. Again, thanks for coming on.

Helena Zhang

>> Thank you.

>> Appreciate it. Thank you. All right. All right, we're here talking to all the entrepreneurs. We've got investors coming in. Also, senior members in the community are coming together too. So it really is a mixed community. Young, old coming together. Systems people here, they think about consequences, first principles, second order consequences. It's a systems thinking revolution that's happening with AI. And again, we're here in Maui with a beautiful background, sun setting. Thanks for watching.