 and welcome back to the Okta Workforce Identity Developer podcast. Today, I'm joined by the co-founders of the Wellness Startup, facets.one. Natalie has a background in front-end engineering and design, robotics, and engineering management, and Caro brings expertise from physics, data science, and clean energy. Today, we'd like to share the story of how they won first place in last year's Okta AI and Identity Hapathon. So welcome, Caro and Natalie. Thank you so much for having us. Yeah, I'm glad you're here. So to begin at the beginning, how did you meet? How did you end up working together? Natalie and I met, I think we've met each other for four years, but initially, we were both volunteering at Girls in Tech, an on-profit organization that seeks to help more women bring more technical side to the space. And then I was the commanding director then. Natalie was in charge of the technical events, so we had to work very close together for over, I would say, like close to three years. And then, yeah, I think eventually, I left. I changed to a different organization, but the relationship stayed there, and we will always discuss about the project. Natalie started because she was very passionate about facets, and eventually she's like, well, we already worked together for a while, so why don't you join the company and let's work together on this? That's basically how we ended up here. Carolina is very detail-oriented, and I feel like I'm a generalist, so I felt like we had good skill sets, and I knew we worked really well together, and I love having her as a friend. Yeah, I'd say the networking goals of those organizations definitely paid off now that you're winning hackathons together. So, Natalie, I understand that facets.one is kind of your brainchild. Could you tell me about the history of the app? So, originally, I broke out of a relationship during COVID and I realized finding someone to be with for your life partner is extremely hard, and I felt like there was no wellness aspect with dating, and that's kind of originally how we started. We thought we were going to be a dating app with wellness, but as with most startups, your first idea is not necessarily the one you end up going with, but as we continued through that line, we realized being a dating app would be too hard to do, especially since there are so many dating apps who already have volume, so we decided to focus more on the aspect we really cared about, which was the wellness aspect, and so that's how we ended up working together and then, hence, finding that hackathon to help build out pieces of our app. So, you were already a pretty well-flashed-out idea around how to do wellness with facets. Before you overcame to our hackathon, how did you end up finding and choosing the octahackathon? It took us some time to start the exploration in the face of the different features we wanted to integrate into the app, and then there was a moment when Natalie and I had a discussion of, okay, we should maybe join a hackathon and specifically choose them for those hackathons that align with the features that we want to push or prioritize, and then at that time, I think there was a lot of discussions internally about the login, and we saw the octahackathon, and I don't know if you want to share maybe about the challenges, because I feel like Natalie was a warrior of this part. Yeah, I think a lot of it is like, even with the original app that we did, I noticed we had hired some consultants to do one login. They didn't do the full validation for it, and so it had me diving deeper to understand login and also learn about OpenID Connect. Hence, the tutorials you guys provided were amazing, really helped to get the concept across, really helped me make sure I had all my box hacked for programming the login for the hackathon. I happened to be on the judging panel for the hackathon, and one thing I noticed across all the different entries is that there were the explicit criteria about need to use AI, need to use octah, and then there were these implicit criteria about guess what's going to fit well with octah's values and priorities, guess what's going to really stand out to the judging team. You showed this excellent understanding of what would stand out, so how do you think you made those connections? I think that was partially luck, and on the fact that we are building something that we're super passionate about, and I think is extremely relevant to our times, mental health and wellness is a huge problem these days. And then the other aspects I think were both very technical, we were both very entrepreneurial, so combining all this together, I think helped us build a good product for the hackathon. Yeah, and having those well-rounded backgrounds in both business and product seems to have really helped you out. Another thing I've noticed as we've talked about the history of the app in a bit more detail when we were preparing is that you seem to be very flexible in your choice of tech step. How do you choose the right tools for the job? Well, I would say we think first, okay, what is the main goal of the feature we want to introduce? We are also constantly joining AI conferences to understand what's the current state of the art out there, and then we're also sort of wary because every day there are new apps coming up. There are also a lot that will maybe disappear in a few months later, so we want to make sure that the tools that we choose are going to be there in the long term, because it's not only that we're learning how to implement them, but also it would be a lot of wasted time if it doesn't lead anywhere, or if it doesn't provide the tools that we need maybe to scale it in the future, or if they just disappeared. So for us, when we were choosing Octa and we were facing the challenge with the login, we knew that, okay, this is a safe bet, and that's how we decided to participate, to learn and expand in that area, but also it will not be wasted time. And as a bonus, we ended up winning, which we did. We're very grateful for that, the whole experience, so that was even better. Yeah, absolutely. And that's, I think, key, especially with AI and the tech landscape changing so rapidly, just making sure you're not building on a foundation that's about to vanish. How was facet.one handling identity before you enhanced it with this hackathon project? We had done, actually, I think just login and password first, and then we ended up switching to one-click logins, which we had a consultant build out, and after I checked the code, realized it wasn't fully implemented. And so, for this hackathon, when we were focusing, we looked at Octa and then learned about OpenID Connect, and then that opened up other avenues of potential thoughts we had on how our product could work with organizations. What are your thoughts about leveraging the features that you get by implementing OIDC and SCIM? What does that open up for you as the product grows? We've discussed different ways, so, or in the future, we see, okay, how can we monetize our product essentially helps the user to become self-aware in different areas, and we're implementing the ACT, which is acceptance and commitment therapy through focused activities. But then these will have a certain limitation of, like, okay, what can we not provide needs to come from an expert therapist. So, if we give therapists the access to, you know, manage the patient's role of our app so they can either provide specific activities or see what they're doing, so their sessions are more fruitful, I think that is one route for the SCIM aspect. Then, I think Natalie had other ideas that incorporated maybe offering a service, say, a university or, like, work benefits that they could say, hey, here is the account because you're already part of organization, and then they just extend that benefit to the final user. Having the control over what you're sharing with these protocols could be really key saying not only does this person get an account, but are they using it without necessarily having to share how they're using it? Totally. One thing that I noticed when assessing the submission is that facets.one seems to really play for the strengths of AI. For instance, if it guesses wrong when it's summarizing a journal entry and says, are you saying you basically feel like this? And then you say, no, that's not at all how I'm feeling. It's actually like that. That kind of is a success in itself. So, what thought process went into scoping how AI fits and making sure that mistakes made by AI wouldn't be potentially really problematic? I mean, I think we always knew, like, hey, with anything around mental health, we always want to be weary. And so we definitely wanted to take an approach of more of a support, a coaching, giving direction and helping the person just understand themselves better. And so I think in the process of using AI to leverage that and just having them validate their emotions, we know through studies that that in itself helps a person feel better. We just felt like it would work really well. There is also a term of like, are you using AI to enhance or replace? And I think we're trying to enhance the user to maybe do these activities or micro journaling that will help them become aware instead of, oh, we're trying to replace the therapist, which is not the case. So for our specific application, we are just like making sure that by having AI to say, hey, you can move very far away from this activity that you're trying to accomplish. And the goals are very clear. And through coaching, like Natalie is saying, having this question that will always pull you back to like, hey, focus and center on this. That is what we're trying to do here. Sounds like you're casting AI as more of an assistant than as a leader necessarily, or it's doing more coaching than therapy or instruction or even mentoring through this process. And with Facet.1, it's really the whole life cycle of starting out building an app through winning a hackathon, through what comes next. What have you learned about building apps? What advice would you give to people who are just starting out with their idea? Go with whatever is the simplest. We have asked product people and they were all like, go iOS. And that process takes way too long. There's so many restrictions with Apple. And at the beginning, you're just trying to test your idea. And so getting to that as quickly as possible is my recommendation. Don't do anything fancy and really stick with simple development first. I will also say, accept that you're going to frustrate it, that you're going to feel like, oh, in your head, you know, things are going at a speedlight. But also, you know, reality and like when you're doing investigation or serving people, et cetera, it's a continuous learning approach. So we've also pivoted multiple times because we have to integrate the learning and also keep our features minimal, because otherwise, it will be an endless journey. What do you think about deciding when it's time to pivot and which features it's most important to keep with you versus which ones it's time to let go of? I think that's always a hard, but I think with some of it is just like feedback by constantly talking to others, getting feedback, seeing if you're getting any traction on it. Doing the combination of that, I think a lot of it is also being realistic about the big picture and all the information you're getting, kind of trying to see where does this fit in that puzzle. We also, as startup entrepreneurs, you're wearing multiple hats. And one of them is, hey, we want to apply to accelerators or different programs like pre-seed, VC funding, et cetera, et cetera. So while you're doing all these applications, you're learning, okay, what are they looking for? And then a lot of times they give you feedback, other times not. Sometimes we're in a pitch competition and then we get people that come close to us and say, hey, I found this part very interesting. And that part might not be our main feature, right? So you need to be able to be open and say, oh, what are people reacting to? And then continue and focus on that. Because if one person say it's like, okay, it's always, every time you talk to someone, you might get a different answer. But if you're starting getting it twice or three times or more, then it's, okay, there's a clear signal here that maybe it's worth pursuing. One really exciting thing as a hackathon judge is looking at how far does this hackathon project have before it could potentially go to market. And you'll see somewhere it's like, oh, a core concept of this would need to change to get it secure. Or a core concept of this would need to change to get it safe with AI. We didn't see those glaring major changes because of how you implemented AI and identity in facets.one. But what do you think is standing between you and being ready to release so far with it? You know, as an MVP or just getting our product out to be able to test, I do think it is quite, you know, ready in the sense that if we make sure everything's secure, logging is working properly, we can probably put it out there for testing already. Fortune right now, nothing's hooked up to any database. So you will have to download your content. So I think that could be something we hook up next. So then there's longevity in the things you built within our hackathon. But the value that our features bring, like just by doing the activities are already there. So I also think that recognizing what is the stage we're at. So if we're saying, hey, this is our beta product, we're testing this specific feature, extend that knowledge to our beta testers. And then we're still learning. We're still in the product fit process. And then once we say, hey, from all this work we've done, this is the strongest signal, and we can put it out in a, you know, like, or for method conception, I think then we have to be 100% sure to to say, hey, it's going to be really safe in terms of the identity, in terms of encryption and all other data. But I think it's going through steps. And we are very serious about making sure it's safe, right? Yeah, absolutely. Anytime you're working with personal health information or even just private things, like once you're in it, it becomes so important. Can you share some examples of ways that you've made it safer already as you've been working with it? I know you've added identity and robust login through the hackathon. So that's a key one. But are there others? Our backend has, at least we had built out prior to this hackathon, had off guards and whatnot or encryption. So if you were dealing with passwords, it would actually in the hackathon, we did that, it would encrypt it and decrypt it to validate your passwords. So those were some, some safeguards we put in. And then also, I think just in general, we at least didn't, from what I remember when we were building, we tried to keep the identity and information separate from our data, like we have different tables and whatnot. There are other things we discuss incorporating. So right now, in terms of journaling, specifically, not only say like a lot of apps that do journaling, they can provide like a backup, for example, but then it's not usable. So by looking at other apps, for example, that say, hey, actually, we don't keep your data, either you use it on your own Google Drive, or you just, you just have the program in your own computer, and you can download your data and it's just there for you as locally. So those are like the two roads that we've been also discussing lately to make sure that maybe a part of the process is given to, for example, a larger company, or it's just like it just slips into your machine and no one will have access to it. I really personally appreciate the approach of letting people choose where it goes, or you kind of get this with using OIDC to do social login from a provider where someone already has all their identity to make the individual security posture no worse for having used their app than it would be if they didn't. Like if I would be keeping a journal in, let's say, Google Drive, or an app that backs up to Google Drive, it may not be perfect, but it's no worse. Exactly. And also, it gives us time to focus on how can we make your experience in journaling better? And then when it's time, we can focus on, okay, do we want to now take this new package into our whole like the software creation? Yeah, for sure. And that's a theme that shows up over and over. What is worth building ourselves? What can we really add value by building ourselves versus what is it better to have someone else an expert, a specialist handle? You also mentioned something interesting about MVPs. So how big do you think an MVP really needs to be? It probably needs to be a lot smaller than you think. That was definitely a learning for us. Even in trying to make things small, I think we realized we can always go smaller. Just to compliment Natalie is when you think of your MVP, you should like cut it off to the point where like what is the actual actual MVP like it's probably going to be like a quarter of what you initially thought was your MVP. Because I think like as humans, we have all these connections and patterns of how everything is working together. But sometimes what you need to test is just like this small part. I don't know, at least for me, my perception of what I thought the MVP was to now that we have this actual MVP like, okay, this is the first feature out that we want to test. It really changed a lot of your time. So if it's taking too long, and if we were saying like, oh, doing the hackathons, we've learned that sometimes doing it as a mock-up, maybe on a web development instead of for an app like Apple, etc. It just like changes drastically the time that you can get and see and test. So I would say if it's taking too long from what you initially thought it would take, then you're not doing the MVP. And you need to like step back maybe and think about, okay, what do I need to do to see this prototype running in two weeks from now, for example. That sounds almost like a mindset shift from an integration testing kind of mindset to a unit testing kind of mindset, but with product features instead of just with pieces of code. Yeah. And so after all this, what's next for facets? I mean, we would like to release the features that we've developed and start beta testing them. Cool. And what's the pitch for who might want to beta test? Who is the ideal user at this point? At the moment, we are still targeting singles in the dating space. But our wellness features can help with anyone. So if you're trying to gain more self-awareness, be happier. You can sign up for our beta. Amazing. And where can we go do that? Facets.one. Yeah. So if you'd like to hear more about the other winners, the other apps that placed in the hackathon, you can check out the blog post linked in the description below. And you can visit facets.one to join the waitlist. And as always, we would love to hear about you and hear about your thoughts in the comments below. So thank you for listening.