Name: Codex Creator Challenge AI Showcase
Uploaded: 2026-05-20T18:31:33.603Z
Duration: 59 min 12 s
Description: Codex Creator Challenge AI Showcase

Transcript for "Codex Creator Challenge AI Showcase": Alright. Hi, everybody. Thank you for joining us today. We're gonna give everyone a minute to join. My name is Randy Tarnowski. I'm a director of research and economist at Handshake. And as we wait, I actually wanna make this a little interactive. We have a poll. So head over to the poll, tab. You should see a little pop up. We'd like to know kind of your experience building with AI. I'll read out the results in a bit, but, go over to the poll tab, and we'll start soon. Alright. So I think we're ready to get started. My name is Randy Tarnowski, director of research and economist at Handshake, and we are so excited today to welcome you to the Codex Creator Challenge AI Showcase. For those of you who might not be familiar, the Codex Creator Challenge is a skill building competition we put on with the team at OpenAI. The goal is to give students free access to try codex and to try to build cool projects using AI. This is to help students build AI skills and really showcase, a lot of the cool things that are coming out, the creativity that's emerging. In a few moments, I'll be joined on stage by Leah Belsky, VP of Education at OpenAI. But first, I want to put all this into context with data from our recent report. You know, like, AI is a very, very big topic, very wide ranging, and there's a lot of real important conversations we're having around workforce impact, adoption, environmental impact, and more. We hear all those questions. There's going to be a lot of topics to cover. What we're really going to be focusing on today is a narrower slice, what's actually happening in the early talent labor market, what students are building, and what employers are hiring for, and where the economy is heading. So that's the lens for the next hour, and I'm really excited to get into it. Great. So, this first slide just gets at a little bit of the adoption that we've seen in a very short time on Handshake. It's interesting because AI is the first general purpose technology that was introduced to, if your students in the audience, you as a freshman perhaps, a tenured professor, and a CEO all at once. If you think of the Internet, if you think of the personal computer, all these things rolled out kind of slowly. But, November 2022, ChatGPT was released and this just grew massively. So a lot of our research and what we're thinking about lately is how do we see this on the Handshake Network and how are employers hiring and how are students learning? So you see on this chart, even just looking at the classes of 2024, forty five percent of students graduating that year had said they had, no experience using AI. Like, even it had been out for two years, and you saw that growth. But even then, from the class of 2025 all the way to today, you're seeing only 15% have not used AI. 85% of seniors today are using AI in some capacity, with more than a third using it daily or more than daily. So just TLDR, it's grown massively, and we're still all wrapping our heads around that. On the next slide, you can look at how employers have actually experienced that. So on the employer end, if you look at the top left there, you're seeing this, like, chart where it goes here, and there's a real takeoff in 2025. That is every job description on Handshake that mentions or talks about AI. So what we found was that it's around 5% today of full time job postings are talking about AI in some capacity. That's twice as many as the year before, and it differs vastly by internships. So if you look at internships, one out of every 10 internships are talking about AI now, and you're seeing in places like financial services, but especially tech, about a third of job postings are talking about AI. Financial services, this is more than three x over a year ago. So it really feels like today, employers are increasingly seeking out AI talent. And if we double click into that, what are the job descriptions talking about with AI? These are just a sample of some of the roles that we're seeing emerge. So things like an AI systems engineer that's expecting, new entry level grads to touch on NLP, natural language processing, knowledge representation, front end visualizations, doing a lot of different tasks. You have things like the junior forward deployed engineer, which is a new role that has really emerged here, talking about making solutions work for specific clients, bridging the gap between what potentially AI can do and then workflows for current organizations, which I think the employers on the in the audience probably are thinking a lot about right now. What do we do with this new tech, and how do we integrate it into things we personally do? Then you have other roles, such as content creative directors that use AI or tracking trends and talking about it. You have compliance analysts who are thinking about how do we use these tools in ways that actually make sense for our organization. And then people leveraging AI and design. So just a lot of different ways this is bubbling up, and we're going to continue tracking this. And then finally, I kind of like I personally love this slide. This is for those of you who don't have tryptophobia, it's like that little those little dots, like, don't like those? I'm sorry. But every one of these dots in these circles, they represent a resume talking about AI. So in 2022, and we're thinking we're talking like hundreds of thousands of resumes. In 2022, seniors then before ChatGPT launched, just a series of small dots. That has 9x grown when we look at seniors today. The interesting part of this is it's not just computer science majors. Two thirds of those that are mentioning or talking about that in their resumes are not computer science majors. And interestingly as well, these are not individuals who are learning it in the classroom. What we found is that actually increasingly, students are building tools like we'll see today. They're building things on their own. They're doing it in clubs and internships and just experimentation. And I think a lot of that creativity, is missed sometimes when we talk about that. So I think today is a great opportunity to dive into that. And so with that, perfect transition into welcoming Leah Belsky, OpenAI's VP of Education, for a conversation about all of it. So, Leah, welcome. Awesome, Randy. It's so nice to meet you, and I'm so excited to be here with this entire community to see what the award winners built and just to spend time with you all. I know. I love that. I mean, I think there's been so much debate and conversation about AI and what that means for hiring, but I know you've had a ton of time on college campuses and talking to students. Like, what makes you really optimistic about what this generation is bringing and for especially the students on the call? So first of all, I feel so connected to this generation of students. As you said, the the students graduating college today are the first ones to have experienced AI through all four years. When I first jumped into this role about two years ago, we found that over 40% of the users of Chatibiti were 24, and over 20% of all, questions on the platform were connected to learning. What that signaled to me is that students were leading, in grassroots AI adoption. And I think these students in particular, this strategy has gone from a lot. You've had to contend with new definitions of cheating and AI detectors concerns about the jobs and workforce and just a huge amount of uncertainty. But there's a couple things that I've observed that have made me just incredibly optimistic. The first is that, you know, when I see students on campuses using AI, contrary to sort of popular thinking that maybe AI is about cheating and outsourcing work, I see that AI is just amplifying ambition. It seems to have locked a new sense of agency and possibility for so many students who would like to build, but I'd say to this date thought they needed to hire an engineer or, you know, network or find a bunch of money to be able to build and create. And that's just a a huge inspiration. But more importantly, I think that sense and ability to build and feel a high sense of agency is critical for the future. You know, none of us really know what the future workforce will look like. We don't we wanna sort of be prepared to both enter the jobs that exist now but also create new ones. And so, that agency that I think AI has unlocked for so many just makes me hugely optimistic. The second thing that makes me optimistic is just that, you know, there is nobody more concerned with sort of cognitive offloading or the negative impacts of AI than students themselves. Certainly, parents are concerned and policymakers are concerned or educators are are concerned. But when I talk to students, you know, these are students who have gone through the ups and downs of social media. They've gone through COVID and and online education, and they are conscious in the same way that they're conscious about nutrition or health. They're like, I need to understand what it means for my brain to learn. How much friction does the brain need to undertake? What does it mean when I cram for a test versus learn something in ways that will sort of stay with me over time? And I would say that level of conscientiousness and concern for for learning and true development also makes me deeply optimistic. And then the last thing, you know, when I talk to students on campuses, I just you see just this burst of confidence for so many people. One person who really made an impression on me when we first, started on our journey in education at OpenAI, we launched these things called Chashqiu Labs, bringing together some of the PowerU sort of students. And there was this woman, Maggie Wong, who was a computer science student at Princeton, and she told this story of, you know, being in the kind of weed out early computer science course. And she would go to her TA's office, and she would sit there for hours and hours and hours. And then she suddenly realized she could actually ask a lot of the questions she had, to Chatibiti, and she didn't have to wait in line for her TA, and she didn't have to sort of marshal the confidence to get into those office hours. And she just said, like, quite plainly, Chatibiti has given me the confidence that I can learn anything. And I think that sentiment, just makes me excited because it means it has the potential to unlock, sort of learning and confidence in so many people. So these are the things that sort of would do acknowledgment of all of the the risks and the concerns around AI, and I wrestle with them every single day within our education team. I think these are the things that really give me optimism and excitement about the future. Okay. That is a that's a healthy healthy amount of of optimism too. And I really I think that resonates to me with me that, like, whenever I hear there's a simplistic idea of someone you like, oh, this group or this generation uses it, like, in this way that's not right. It really reduces, I think, a whole group down to a very simplistic and I I have a similar finding with, like, when we talk to students, they're, like, very discerning and they're very thoughtful about how or when. And I think we're all kind of growing and evolving with that together, so I'm glad that that is like the emergent kind of optimism point. Okay. So another piece that I think is pretty interesting here, and you kind of mentioned the COVID piece and, like, growing this generation. One part of the research we found that I found most compelling is that there's this 30 percentage point gaps between students that feel like they learned how to use AI in their their higher education experience and then 30% higher say, I'm gonna need to learn this from, like, for employers or in the the workplace. And that gap is so interesting to me because I I feel like that's what we're seeing with the creativity that students are exploring. So I'm curious your take on that, but also for the students watching, you know, what's an underrated way that students can actually showcase this to employers? So So it's a great question, Randy. And I think, like, what you're hitting on there is a big need and change for the future of education. You know, for so long, things like makerspaces or entrepreneurship classes or or builder activities were a sideshow at education. They were left to a separate building or to an extracurricular activity. Part of that was because the means to create and build was scarce, and it took so much effort for an entrepreneur to get from zero to one that there was sense it was only a few special people who could manage to do that, but that's not true. And I think we're in a moment where more people are realizing that sort of building and creating needs to become become part of the core of education, something that everyone can experience and sort of and and and develop skills for. But, you know, given where we are, one, I super appreciate, like, the lift it takes for people like those who joined this codex challenge to be skilling yourself for what you need think the workforce needs and building projects while also getting what sort of institutions traditionally think you need to get in order to get your degree. And I would sort of give three simple pieces of advice. Like, one, you know, the the workforce and Randy knows this this information best, but I think the workforce is changing such that, like, more people are fewer people are being taken on sorry. Fewer people are taking on tasks that are given to them, and more people have to actually identify problems or projects or needs in organizations. And so I think to be prepared for the workforce, you need to increasingly show not just that you have AI skills or not just that you have an ability to learn and that you're an active learner, but that you can take on a challenge and build something and solve it. Last week, we we launched a program called Chat GPT Futures where we highlighted 26 students, that had built some amazing things with AI, and and the projects were sort of basic. They were mapping food banks. They were looking at new models for doing financial analysis. They were rethinking sort of approaches in discovering stars that that wasn't wasn't so basic. But that's what I that's what I tell people. Figure out needs or problems that you either think the businesses that you're applying to have or that you see in your own community and take the process of building and create something to solve those problems. The other thing I would say is it's now more easy than ever to start an organization. I was recently hanging out with, a a leading entrepreneur in Israel whose whose son just built a political organization, and he sort of put out an idea and then gathered people on WhatsApp. And when he was called the government to show a minister what he built, he sort of scrolled through WhatsApp, and he said, look at all these people who are who are connected to my organization and this idea. And what was striking about it is he had skills on social media. He had an ability to sort of build the following in ways that many of the people in the workforce or in the government today didn't. So it's just another beautiful example of sort of how meaningful it can be to those who are looking to hire when they encounter young people who have AI skills, who are learners, and have shown ability to create and build. Love that. Yeah. And I I think that kinda leads into this other one that's, I think, a nice tee up, because I think I know we can where we can go with this. But there's this idea that AI, and I I think maybe an antiquated one, that AI skilled means, like, oh, it's computer science major. It's the one that does computer things. But, resume data shows that kind of massive growth, especially outside of computer science. You've talked about things like live coding and lowering the barrier, but, like, what might that look like for folks, not just in engineering, but things like marketing or finance or other, you know, just the vast majority of jobs out there? Yeah. Absolutely. You know, look. The the statistics tell a very clear story. Today, Codex, the the tool that you all use in the hackathon, has over 4,000,000 weekly active users. One of the fastest growing segments are noncoding general knowledge workers. And within that segment, one of the fastest growing segments is actually students. And I think that's reflective of sort of the appetite we saw around AI early and with chat GBT. But, you know, I also interact regularly with enterprise teams who are bringing AI into specific job functions, and you see very clear patterns of an area of an expertise combined with specific tasks and a role combined with AI that is used to sort of build those solutions. So to give you an example, like, our CFO recently gave a presentation with the finance team, and she showed the way her team was using AI to do new financial models. And last week, you saw the release of a new Excel powered AI tool. She showed how it was being used to do due diligence with investors to sort of bring together a body of knowledge and streamline those investor conversations. I recently hired someone, and I hired him as a contractor, actually, to take care of most of our marketing channels. And he came in, not a coder at all, and he built what he calls the derivative content machine. So the machine that would allow him to create one piece of content and then modify it for LinkedIn, for for Twitter, for all of the social handles, and then bring back all the analytics. And it was built just on codecs with a set of skills and an overlying, application. So I think that's the real power of these tools. For sure, they accelerate coding and and engineering, but what I'm seeing every day is sort of young people who are realizing that now they can they can be builders. They don't need to be engineers in the same way to to realize some of the dreams and projects they want to realize. Love that. And what a perfect transition. So thank you, Leah, for all that, all your insights there. I'm sure we could probably go on all day about this, but, now I think we're gonna transition into, showcasing some of these great projects. So. I can't yeah. Thank you so much, Leah. Okay. So now, if you have drum roll, you know, drums, snare drums, and all, feel free to, build that up. This is the moment where we get to see some of these projects in action. It's worth noting that during this project, we received over 1,500 project submissions. So what you're seeing here, you're going to see three of the most creative, impactful projects, but the diversity and the extent of the projects that were submitted are really impressive. So if you're at home, I'd give you full permission to clap at your desks and get excited about this. But let's take a moment and recognize everyone involved. But up first, you're going to hear from our first place challenge winner, Obina Wanchukwu. So Obina, congratulations and welcome to the AI Showcase. Can you tell us a little bit more about your project? Yeah. So, yeah, my name is Albina. I just graduated Georgia Tech for CS, and I wanna talk a little bit about Tracecode. So Tracecode is an interactive platform for students to learn coding algorithms. So most technical, most technical interviews in this field consists of implementing an algorithm that you've learned, and many, many people use platforms like LeetCode to memorize solutions. But I find myself as a visual learner, memorizing hundreds of algorithms doesn't actually help me learn what the algorithm algorithm is doing. I'm just, like, memorizing what I wrote. And I really internalize best when I see an algorithm visually work, like, through state changes, and trace code is a product of that idea. So it's a platform that doesn't just run code, but it can visualize, though you can see exactly where it's going right or wrong. And I can dive straight into the demo if I can if I can screen share. Do yeah. So it should be right up on the oh, yeah. There you go. Yes. Yeah. So, can you guys do my screen share? We do. Okay. Perfect. Perfect. So today, I wanna show off two features. So I wanna show off codelessness and visualizations. So as you can see, this is a binary search problem. Right? And I if you have keen eyes in this audience, you probably can see what's wrong with it. But let's say I made the solution. I'm gonna click compile. And as you can see, I don't pass all the tests. Right? And, you know, I'm like, wow. What am I doing wrong? Right? So I go to this button up here, go to this, and it immediately tells me the binary search loop may not be starting from the full search interval. So that has me thinking here. And the issue is that I'm starting out of the range of the array. Right? So minus one, the issue goes away. If I compile, then I pass all my tests. That's code assist. It's a way for you to unblock yourself if you're stuck. And usually with, like, off by one issues, those happens to me all the time. And then for visualizations, if I step through the visualization, you'll see as I'm going in, the variables are actually being projected onto the array so you can see what the algorithm is doing. Like, as the left pointer moves, the mid pointer is moving. And yeah. So both of these features were powered by a semantic engine that actually understands your code. So in this case, they detected that we were trying to do binary search, but we were off on the right pointer, and Codasys, passed down that hint. And the engine can also detect workflows like graph, trees, and stacks, all common algorithm types, and the visualizer uses that as proof to render the data structure. So if you have, like, a graph problem, it'll render a graph with edges. And this is all powered with the engine without any LOM calls for really low latency. And all of this was accelerated using codex. I was creating this solo, but if I could call anything my partner in this challenge, it would have been codecs. It allowed me to start from a blank project and grow to grow grow to what it is today. And I could go on for hours talking about how I use codecs and the hacks for my workflow. But with these tools, you really just have to try them for yourself and see what works for you. And I just want to end by saying thank you to Handshake and OpenAI for making this challenge and inspiring so many people to build innovative products. And that's super impressive, Obina. I actually wanna if you don't mind, I'd love to ask a follow-up on this. Yeah. First. Also, it's just impressive. It's like a very meta level thing. I think the UI on this, it just looks so professional. I guess, really, where do you see this going after, like, do you see yourself continuing to grow this? Yeah. So during this challenge, I implemented a bunch of features. But after this challenge I mean, I've already had, since this challenge ended, a bunch of features that I've been working on, like more language support, better visualization. So, yeah, I see this continuing until I'm until I'm satisfied. Yes. Okay. That's great. I mean, like and you're saying there's no LLM calls. This is all actually like, this is just really responsive in the app, these corrections and whatnot? Yeah. That was the hardest part because, I mean, it's easy to just, you know, LLM call what's wrong with my code. Yeah. But the hardest part was trying to engineer a system, and it was mainly with codecs. Like, I'm gonna say today, like, if I tried to do it, it would have taken me years. And I did the system and got it working enough within, like, thirty days. So, yeah, I kind of had the idea for how the system would work, but Codex was the one that, like, really drove the implementation and even pointed out, like, hey. This might not work. Or I would point out, hey. What are you doing? This is not what we're you know, this this won't work. So it was really that back and forth that made the engine into what it is and powers the system. This is, yeah, so impressive, and I think, like, it's why we probably need to come up with a better term than vibe coding because I feel like vibe coding suggests you're just like, chill. This is very robust and very deep. So congratulations on this project. Yeah. Super impressive. Thank you. Thank you. I've been hearing people say agentic engineering. Maybe that's the next term. Alright. There we go. Alright. Well, thank you so much, Olivia. I think we're gonna showcase the next one. Okay. So next up, we have. Wei Ying Chung, our third place winner. So welcome to the stage, Wei Ying. Please tell us more about your project, who you are, everything. Oh, you're on mute. Feel like it's the bottom left. You should be able to unmute. Oh, playing. I I still can't hear you. There's a believe the bottom left is the mic. Oh, can you see my screen now? I can. Yes. Hi. Okay. I guess I mute myself. Hi. My name is Huynh Zhong. I'm from Arizona State University. My major is graphic information technology with a focus on UX design. So, my project is called Where Dragons Dwell. It's an interactive atlas of dragon myths across 14 cultures. I chose this topic because, I have limited coding skills, so I chose to focus on UX design. And that's why I created this cultural exploration website. The similar framework sorry. The similar framework could also support other cultural topics like folklore, mythology, and anthropology, history, and the unknown, and those are for educational and academic use. I also included, casual users. So for better experience, I narrow it down to MVP scope because I know people are familiar with dragons. They saw dragons in movies, novels, folktale, video games, etcetera. Different cultures interpret dragons differently. And with this website, you can compare and explore the dragon culture. I make sure it's like an immersive exhibition and make sure it's visual appealing. Through the process, I use chat GPT, help me in the creative and research process, creating visual elements, draft, and prompts. Codex implemented almost all of the feature functions, including timeline, comparison, missed card, and draw card. After that, the responsive design iterations and debugging are built with Codex. Right now, let me show you quickly of the live site. I guess I have to reshare the screen. Yeah. I'm. excited. Can you see it now? Yes. Okay. The for the, first, first time users, you can come here and hit how to explore and see the instructions. After you read through it, you can come here and see the timeline feature and click each either one, and you can see different period. Dragons appeared on the map. Yeah. They they have different different, history period. And then, if you hover your mouse over the map, it will show the show the dragon's illustrations, and you can click either one you are interested in. Let let's say this one. And then the interface will show you the, detail information of the dragon, and you can read through it. The information is very detailed. And then if you are, curious about other other dragons compared to this one, you can use the drop down menu here and choose e choose the the one you like the most. Let's say this one. I don't know what that is. Let's see. Compare. And then, the the interface show you the two chosen dragon comparison. And you can read through the comparison and see if that answers your questions. If no, that's okay. Because if if you have open questions, you can also use the AI oracle here. Let's ask this one. The model is GPT four mini, and this works very well. You can see I have preset, topics and questions. You can use, anyone anyway. Yeah. So, I hit this one and then hit send. While you are waiting for the oracle. Yeah. Yeah. Yeah. Yeah. It's very quickly. And then, if you have your sorry. If you have your own questions, you are welcome to type here. And I'll skip this, for now because oh, okay. Yeah. Well, I'd love to ask you a question about this, Lee. This yeah. And then yeah. Go on. Oh, it it does have have boundaries because I don't want it to be unrestricted. It's not a general chatbot. And then let me close this. Close. And let's use the draw today's dragon card. This is my favorite part because I was inspired by Japanese capsule toy. And, everyone can draw only one time per day for the card. Let me draw. Oh, I like this one. This is a this is a western dragon. If you like it, you you are welcome to open the dragon entry and see more information about this dragon. And if you want to collect it, you can print this museum card. And this is the preview page. If it is too small, you can zoom in it and see more detailed information, and then close it. If you like it, just print it out, and there you Wow. go. That that's it. So in. the audience, if you. want today's card, you have to you have to go and look at it. You you have to come back tomorrow to draw another card. Yeah. That's so cool. are total will be 14 cards. Yeah. Well, And have I have drawn by random. yeah, and I have I have so many questions about this, but I I sadly. Mhmm. only have time for one. So it's gonna be smart. I think it's so clear that your design background comes through here. It's just amazing what you're able to do here. This looks so great. I guess, was there something you learned about specifically the content and the dragons from from building this that you really surprised you or anything? You mean the process for the. creative process? or even just discovering more about this topic. Like, did you, hear more about, like yeah. I guess I didn't mention that my background is Chinese literature before I came to United States. Okay. Okay. Yeah. So, You're able to integrate all of that. That's so cool. so the topic, I I'm kind of familiar with that. Actually, my background is more cultural, like, culture and literature. Yeah. Oh my god. So. amazing. Thank you so much, Lee Ying. I love hearing the backstory and how this came out, and I will be collecting this card. So that's great. Uh-huh. Okay. Thanks. I'd like to welcome our final, creator today, Andrew Bybee. He was a finalist and is a creative share out as well. Andrew, thank you so much for joining us today. Would love as, Huiyang is transitioning out, Andrew would love to hear about your, experience and kind of telling us more about your experience, with this challenge. Oh, I think you're muted, Andrew. Yeah. Can you hear me? I can. Yeah. Alright. Perfect. Hello, everyone. My name is, Andrew Bybee, and I built this application called fix my shift. Pretty straightforward. It catches payroll mistakes before they cost you. So before I tell you about the project itself, I need to tell you sort of, like, where it started. So this is where it started. This is my restaurant. Well, not my restaurant, but it's the restaurant that I work at. And this oh, sorry. This is Imo. She is the heart and soul of every recipe we make at the restaurant. Every dish that walks out of the cat that kitchen is because of her. As much as we love her, we wish she would remember to clock out. And I used to think that, you know, it was just a small thing until we started looking at the data. That is me. My name is Andrew Bybee. Like I said, I work at Krispy and Barbecue, part of the sales team. And my job is to make the restaurant as profitable as possible. Every pay period, my boss would spend hours just manually going through everything. A single missed clock out means hours that weren't worked, and, you know, that's that's real money. If we do some simple math here. Right? So we have one email here, and, you know, that's pretty manageable. We could fix that. But if we multiply that across 18 employees and every two weeks, you're looking at real mistakes every single cycle. On top of that, our label was running about, like, 30% above 30%, which is the threshold where restaurants stops being profitable. So most restaurants just take that cost. Nobody catches it before payroll finishes, and that's why I built Fix My Shift. I'll be honest. I'm not a developer by any means. I'm just a art student at Knowman. This whole app was built between shifts, sitting in a booth in a restaurant with my laptop. And, yeah, Fix My Shift uses live time card data and live purchasing data integrated directly with your POS and scheduling system, to make sure your restaurant is running, not just efficiently, but profitably. It catches missed clockouts the moment they happen. It surfaces overstaffed shifts before they blow your labor budget. And overall, it gives you one clean place to see everything before payroll finishes. I know that's a lot of talking, so I'll switch over so we can see the actual product. So this here is fix my shift. I just wanna make sure you can see that right. Yep. I believe you can. So if you go into here, this is the website that I made for it. It goes over lightly, like, what it does and the use case. But more importantly, if you sign in here, I could I could actually access my live restaurant data, but we're gonna look at the demo data with all fake numbers and information. So in this one here, I'm gonna skip the tour for now. This is actually what the platform would look like. At the top you can see the pay period that it is, referencing. It's April '16 to the thirtieth. And we know we're above target because our the labor exceeds 30%. And here on the insights panel, we I just put a little blurb about, like, you know, it doesn't just it tells you the whole story of what's happening and why it's above target. And it gives you recommendation. Weekend staffing, driving labor spikes. Friday, running 47% above labor, drop one shift Friday and Saturday. Gives you great recommendations. Over here on the right, it shows you how we did the math. Right? It gives you your total sales, and you take it divided by your payroll, and you get your labor percentage. As simple as that. And so this is a high labor scenario. I'm gonna show you if it was an open shift. If we had some so for some reason, someone had open shifts, this is what I would do. In this example, we have Mia, Jason, and Lily. So Mia here, I could remind her. I would just click this. Oh, let me just give it to her. I can click this here and it will automatically, if you're on your desktop or on your phone, you can, like, text them or you can send the email out to them. Or on your local, like, browser, you can just fix it, on the on the website. So either way, you can do that. Down below here, you can review how you see it. So you see in front of the house, there's one issue. Back of the house, there's two issues. And you can see, okay, who is it? Oh, it's Mia. Mia is highlighted in red. What's, what's in the back of the house? Oh, it's Jason and Lily, and so on and so forth. And you would just resolve these as the shift goes on. And now let's just see just in general what is the healthy pay period looks like. This is what it would look like. It would still give you insights even though there's nothing wrong, but it gives you, you know, everything, front of the house, back of the house balance looks healthy. Our split is forty seven fifty three. No rebalancing required. And once you think everything is good, you can export it to Google Sheets, and it syncs perfectly here. And it would show this, which is, all fake data, like I said. But at the top, you can see here, this is clearly labeled your pay period. It gives you detailed information about what happened here, your front of the house, your back of the house, and it has a master sheet here where it, sort of uploads and logs all the previous information. So it's pretty simple. It's it's as simple as that, but it it helps a lot. So I'm gonna stop my share here. And, I mean, yeah. I mean, that's basically it. I mean, within the first few weeks, my boss used this. We already caught, like, two missed clog outs. Most small restaurants can't afford the enterprise tools, but they have this exact same problems. Fix My Shift gives them access to the kinds of insights that used only used to only exist for, you know, big businesses. And none of this would have existed without codecs. Like, none of it. I don't know how to code. I went from not coding to building this project because codecs isn't just for engineers. It's for people who truly have a problem and wanna build something to solve it. And, you know, emo still is gonna forget to clock out, but we can catch it before it costs us. And so this isn't built from the outside. It's built from within the restaurant from a very real problem that happens on a Tuesday night. So all in all, my name is Andrew Bybee. I'm actively looking for roles where I can keep solving problems like this, and I appreciate your time. Thank you. Man, Andrew, congratulations. That was incredible. I especially appreciate the Photoshop work, the like, I was really impressive. Thank you. Thank you. Appreciate it. It's cool. I I mean, I love all the projects, and I love that this one is really I mean, we talk about we're talking about all these forward deployed engineers and this idea of, like, we need someone that actually understands the problem so they could build something, and I think you did that incredibly. So I guess I have one question kind of, like, if you could say more about, like, kind of the response in the restaurant or when you shared this out and explained it. Like, what was kind of the response that you got from, folks you showed it off to? Well, definitely, they're like, well, Angie, you don't know how to code. How did you make this? They thought I just, like, you know, I purchased something. They're like, so what's it gonna cost us, you know, to use this? You know? A lot of, like, you know, where's the money coming from first? And I told them no. It's like, a lot of this was built, you know, with codex. All this was built with codex, and it's like, it was made. I mean, yeah, not you don't have to buy them. I mean, I could charge them, but, so, yeah, it was it was, they they really like it. They use it a lot every single day. They just use it to manage. Right? I think the biggest issue is a lot of times, we wait till the end of payroll to catch things. You have to retroactively go back. Right? But this it can monitor and remind you during your shift, and just let people know. So it's a very proactive, helpful device and tool that people can use at a restaurant. So yeah. Yeah. I think small businesses and folks like that are just gonna benefit so much, especially from talented people like you bringing in the context and experience. So congratulations. That was such an impressive project. Thank you, Randy. I appreciate it. Yeah. No problem. And then okay. So now we're gonna transition we're transitioning from the talent side a little bit to the employer side. So we're gonna hear from some codex Creator Challenge sponsors who want to share a peek into how they're thinking about AI skills showing up in their early talent programs and maybe share some advice on ways, if you're a student in the audience, what to look out for. So first, we're going to hear from Heli Nayanani, Director of Campus at GEICO. Then we'll hear from, Brandon Maroney, Career Development Program Director at JPMorgan Chase. And finally, from Erin Wozniak, Head of Employer Brand at ZS Associates. So, really excited to talk to the three of you. And, Haley, welcome. How are you doing? Thank you so much, Randy. Appreciate you having me. As Randy mentioned, I'm Haley Nani. I lead early talent recruiting at GEICO, and so happy to be joining you today. A big congratulations to the winners. What incredible projects. I actually just re started rewatching Game of Thrones. So I was looking at the dragon project and definitely think I need to explore that. That was very cool. So wanted to share a little bit more with you today about GEICO and how we're thinking about AI and specifically the opportunities we're creating for early talent at GEICO to explore their AI skills. So at a at GEICO, we invest heavily in early talent through our rotational programs, internships, and various entry level analyst roles like those you see here. Really, whatever your major or interest, there's a path for you at GEICO. And I think you'll find that the theme of today and what you've heard from many of the speakers is that no matter which role you start in, AI will be a part of it. And that's definitely the case at GEICO. I think early talent really has an incredible opportunity right now to make a significant impact from day one. You know, while you may not bring as much experience to the table, if you have the right problem solving, judgment, communication skills, you can now leverage AI to solve problems, automate work, and deliver value for a company in a way that you never used to really be possible, for a new college grad. At GEICO, AI is becoming part of our DNA, and we're creating this space for innovation. So you'll see some of our entry level talent like Nishal, who's part of our TDP technology development program, you know, had said here, we're really giving interns and new grads opportunities to jump in and make a difference, especially with AI. If you wanna go to the next slide, when I say we're embedding AI in our DNA, what I mean is that AI isn't just a tool used in tech, right? It's our new operating model and it's really central to our transformation and our business strategy. So wanted to give you some concrete ways that early talent is using AI at GEICO. You know, we asked some of our early career engineers in our TDP how they've used AI in the past year. And they told us they're using it to debug code faster, analyze root causes of issues more quickly, and optimize their solution. So this really allows them to deliver results and learn at a speed that just wasn't possible a few years ago. And it's not just for the tech roles. Right? It's everybody. So every role is leveraging AI. For example, our claims and our service teams are using AI assistance to work more efficiently, and most importantly, help millions of our customers get answers faster. So as a new hire, I think you'll find AI within everyday workflows. And what that really means is that you'll be able to build AI skills simply by doing. your job at GEICO. Now we know adopting AI across the company really requires everyone to feel comfortable trying new things. And so a lot of that comes down to the culture that you're creating. So we're actively fostering a culture where curiosity and a growth mindset are celebrated. Our leaders are modeling AI first behaviors and their decision making and champion AI projects. So really you can see as early talent role models for how to integrate AI into your own work. And, you know, we also don't wanna just throw everyone into the deep end. We're investing in helping all associates become AI proficient. So we have responsible AI training to make sure you know how to use tools effectively and ethically. We've rolled out Microsoft Copilot to thousands of employees across GEICO, so everyone really has it at their fingertips. We're offering tiered training so you can progress as you become more confident. And then, you know, we've really tried to adopt a learn by doing approach. So we're running prompt a thons, internal innovation challenges where teams can compete to solve real company problems with AI. And then we've created AI champion networks. So these are enthusiastic early adopters in different departments. Constantly upskill, experiment, and never stop learning. So I wanna close with a couple of practical tips you'll see here to stand out as early talent, especially with AI. So first, when it comes to AI, don't just describe the project you've done or what you've built. That is great, but I think the real value comes when you explain what problem you solved and the impact you achieved. So employers like GEICO love to hear how you applied your skills to create value. Second, I would show that you weave AI into your everyday life. So demonstrate that using AI is now just becoming part of the way you approach problems. So for example, maybe you had a task you were doing in an internship and you were able to automate it with AI. Just showing that you're curious and hands on with the tools tells us that you have the mindset we're looking for because we really wanna attract talent that's using AI as a natural part of their toolkit. You know, I think where you are right now as a student, this is such a perfect time to build these AI schools, these AI skills. So as you finish school, launch your career, you know, embrace opportunities like this, embrace opportunities to use AI in the classroom and for projects, entering competitions. It won't just help you stand out in the job market, but you'll be ready to make an impact when you do launch your career. I think we have some some virtual boosts, right, Randy? So if if GEICO's culture sounds like something you wanna be a part of, we'd love for you to join our talent community. You can sign up in our virtual booth, and, really, just good luck to all of you as you launch your careers, put your AI skills to use, and really appreciate the opportunity to share more about AI at GEICO with you today. Amazing. Thank you so much, Haley. I think, so much of this weaved nicely into, I think, the three amazing projects we saw. But I'm sure tons more on the call right now in terms of students thinking about how to maybe kind of connect with the company that is embracing that. So next up, we have, Brandon Maroney from JPMorgan Chase. Brandon, excited to hear from you. We I think we're having some video technical difficulties, so we will certainly hear from you and see your slides. But, welcome to the stage. Oh, I think Brandon's coming. Okay. I think we might actually start with Erin first. So, Erin Wozniak is from ZS Associates. So, Erin, if you, are there, let's get you on stage to voice over some of these slides. Hello. Thank you, Randy. Hello, everyone, out in the audience. I'm so, so, so excited to be here. And the the opportunity to see the projects from the challenge winners and hear also from fellow employers around, what's happening in the market and what we're looking for is also just is very, very inspiring. First, let me give you a peek into what opportunities are available to students and recent grads at ZS. We do offer a range of entry points depending on where you are. So whether you're ready for an internship, you're looking for full time roles across consulting, analytics, and technologies, and other opportunities beyond those areas to explore different paths and and as you grow. What's consistent across all of those is the experience itself. From day one, early talent at ZS are working on real client problems. Whether you're analyzing data, building solutions, shaping recommendations, You're not just observing from the sidelines. This is one of the the key differentiators that we offer in in these experiences. And, we're really proud of the ability to do that, especially in given the exposure to the type of work that you will, you know, be impacting through the clients that we service. This can seem scary. So we also, you know, make sure that you're supported by a very strong mentorship model and, you know, given real ownership early on. That combination, responsibility plus support is is what helps you build confidence quickly and really understand the the impact of your work. And because we're a global firm, you're also learning alongside teams from different backgrounds and perspectives, which really does accelerate both your technical skills and your, you know, softer skills, how you think about solving problems, how you handle conflict, all of those things. But rather than just hear from me, I think the most powerful full perspectives are from the people who've actually gone through it. So what you can see showcased here are just very few testimonials from recent early career joiners, to share what that experience really felt like for them. I will also encourage you and maybe I'll put this in the chat. I don't have it on the slide as soon as I'm done speaking. But for the full stories behind these quick snippets and a bunch of others, I encourage you to follow our Life at ZS blog on zs.com, our LinkedIn company page, and our Insta profile. We are constantly calling on ZSRs to share their stories and understand from the data and the engagement that we get on that, that's really what helps perspective CS ers, you know, learn about the opportunity here and and hearing it from, you know, the people that are already experiencing it. So, shifting gears a little bit, if you wanna go to the next slide, I wanna talk some about, you know, where does AI show up in your work and what additional advice beyond, you know, what you've heard already, you know, can help you in your in your job search. So, at ZS, you know, where does AI actually show up in the work and what does that mean? And the very simple short answer is everywhere. And it's different, you know, depending on your role and also depending on, you know, your aspiration and aptitude and, you know, essentially, all all are welcome. It is for all CSers across all of our global locations, all of our different delivery centers, AI is part of our daily workflows. It helps us move faster, whether that's synthesizing information, structuring thinking, or accelerating how we develop insights. In the more technical roles, that might look like building or integrating AI into analytics and software solutions. In in consulting roles. It often shows up in how you're analyzing complex data, you know, generating a hypothesis, and translating those into clear, actionable recommendations for clients. Just as important as using AI is using it responsibly, understanding where human judgment, data privacy, and ethical considerations come into play. We've invested a ton in this area in responsible AI under our business responsibility and impact efforts. So, you know, that is a a huge piece of the considerations that that come into play. So given all of that, when you're thinking about how to stand out as a candidate, a few things really matter. So this is where the the advice sorta comes in. So first, similar to what you've heard, show your applied experience. Don't just say it. Don't just say you've used AI. Show how you used it to solve a real problem or create impact. That goes back to what Haley was saying too about it's not just about showcasing the the project or the product, but it's the the why and what problem you were solving. I think Andrew did a really good job of that. Second, be transparent. Don't hide your use. Walk the people that you're talking through, you know, through your thinking, how it improved your work. It's it's certainly, you know, something even as basic in the most simple application, even in your in your personal life, shows, you know, and demonstrates an an aptitude. Third, communicating clearly. This probably goes without saying, but being able to really explain the the AI driven outputs in a way that nontechnical people, which, you know, will be a lot of the people that you talk through in your in your job search, in a way that they can understand is incredibly valuable. And, you know, ultimately, Haley kind of referenced this to, you know, we're not looking for AI experts. We're looking for the aptitude and the competency to be curious, to be thoughtful problem solvers, and, you know, those that really know how to use it or learn how to use it to make it better. Love it. Thank you so much, Hailey. Aaron, I'm sorry. Really appreciate it. I believe we can we will push people to the employer boost when they have a chance and be able to dive into all this, but really appreciate your insight. We're gonna try having yeah. Of course. And we're gonna try one more time with, Brendan from JPMorgan Chase. Brendan, are you on the line? Okay. This is the moment of truth. Okay. Well, in that case, we will, forward there will be an employer booth for JPMorgan Chase. Here are some of the slides that Brendan is sharing, related to the early talent program. You can definitely check that out in the employer booth. And a little bit more on some of the ways in which JPMorgan Chase is having AI show up. Software engineering, AI is embedded into their how software engineers are building and deploying technology. AI solutions are being leveraged to scale models, improve data quality. And so the advice here from Brennan is showing how you use AI by highlighting how you use it. I think we got to show a lot of examples of that today, but also validating outputs. Communication skills we hear that again and again, that this is so important in a time in which many of the technical skills are emerging and able to be touched on with AI, that communicating what those are. And we heard a little bit from this from Aaron. Communicating what those are and how it translates is going to be critically important. And then finally, demonstrating big picture thinking by showing how your work solves a problem. I think Andrew, is a great example of how we started with a problem or an opportunity, and then we solved it with this versus finding the problem with AI. So here's some that's some of the advice from JPMorgan. And, thank you so much. That was a very packed day. I have a few other pieces to call out. We want to thank, obviously, all the employer partners here and, of course, all of the creator challenge winners. Be sure to see the employer it should be, like, right above my head. The employer booths on the top. You'll see not only the employers we have today, but also, Salesforce, UBS, L'Oreal, and other employers. So definitely check them out. And I also, before we say goodbye, want to say a few things about ways Handshake is really allowing students to showcase their AI skills. We've seen this demand really grow over time, so we want to create an accessible way to show this off. We launched the AI Showcase Hub, which is a destination where students can learn and experiment, and it doesn't matter your major or your experience level. It's just an opportunity to get things out there. Our vision is that this hub will bring together prompts and tools and practical challenges, so folks can get going quickly and employers can learn about what kind of skills are emerging from candidates. Additionally, we have some cool badging opportunities here. So there's the ability to showcase what you've been building, making this participation in this AI showcase very visible to employers. So if you we've heard from a number of employers today about how, like, showcasing these skills are important, and I think this will be another great opportunity for that. These badges are going to help students stand out and show their experience with AI and allow employers to check out directly what they've built. We're going to continue to invest in this as AI evolves and things grow, but stay tuned. But until then, thank you all so much for showing up, for the guest speakers, sponsors, and codex creators. Big thanks to OpenAI, Lea, and the team, and thank you all for tuning in. And hope to see you again soon.