Tech-Savvy Without the Code: Building a Startup as a Non-Technical Founder
- Stephanie Roulic
- Oct 10
- 42 min read
Every founder dreams about building the next breakout startup—but what if you don’t know how to code? For years, that gap was seen as a dealbreaker. Today, thanks to no-code platforms, automation, and AI-powered tools, the rules have changed. Non-technical founders can move faster, test ideas sooner, and bring products to market without waiting on engineering talent.
At Startup Boston Week 2025, we sat down with four women who’ve done just that—turning bold ideas into real ventures without writing a line of code:
Elizabeth McWhorter - Co-Founder and CEO, ThriveIEP
Sophia Lin - Partner/Founder, Dots Global/ Esquare Lab
Jamie Gong - Co-Founder and CEO, Care Hero
Karen Kelly - Founder & CEO
They shared how they’ve leveraged AI tools, off-the-shelf platforms, and smart partnerships to build and scale companies, while also navigating the challenges of fundraising, credibility, and managing technical projects.
From breaking through the myth that you need a technical co-founder, to learning when to use off-the-shelf tools versus custom builds, to understanding how to pitch investors as a non-technical founder, the panel delivered candid, actionable advice for anyone looking to launch without a computer science degree.
5 Takeaways
1. You don’t need a technical co-founder to get started
The old wisdom was simple: you need a “hacker” and a “hustler” to start a company. But panelists pushed back on that idea. With the rise of tools like Replit, Supabase, and AI-assisted coding platforms, non-technical founders can build an MVP in days, not months. Instead of rushing to find a CTO, consider leaning on fractional experts, advisors, or simply experimenting on your own first.
As Karen Kelly put it: “Today in 2025, you can build these things in a week. You don’t need a technical co-founder to start.”
2. Paid customers > polished product
Anyone can spin up a prototype. The true test is whether someone will pay for it. Whether you’re pitching investors or considering raising capital, traction in the form of paying users matters far more than a sleek demo.
Elizabeth McWhorter emphasized that early-stage credibility comes from proof: “There’s an angel investor who always says, ‘Show me what you built. Build something and then come back to me.’ Now you can do that in a week—and what really matters is whether people pay for it.”
3. Off-the-shelf tools are powerful—use them wisely
Custom software isn’t always the smartest first step. Many successful businesses can run (and even scale) using existing no-code platforms, automation tools, and CRMs. The key is knowing when off-the-shelf tools are “good enough” and when it’s time to invest in something custom.
Sophia Lin, who scaled a real estate venture to $15M in assets without writing a line of code, summed it up: “If you’re just starting out, off-the-shelf tools are good enough. You don’t need to custom-build until you’ve verified product-market fit.”
4. AI is a teammate—not a magic wand
The panelists were bullish on AI’s ability to accelerate product development, but they warned against blind trust. Treat AI like a junior team member: it can get you 80% of the way there, but you need to double-check its work, ask follow-up questions, and sometimes even have different AI tools critique each other.
As Elizabeth put it: “Assume that AI tools will lie to you—because they will. But if you treat them like a new hire, you’ll learn how to get the best results.”
5. Don’t sell “AI.” Sell the benefit.
Customers don’t care that your product uses AI—they care that it saves them time, money, or headaches. In fact, leading with “AI” may even hurt your pitch, since many consumers are still skeptical.
Karen Kelly cautioned: “Build a product that’s made better by AI—but don’t make AI the thing your customers are interacting with. Power your product with it, don’t market it as the product itself.”
Final Thought
This panel made one thing clear: being a non-technical founder in 2025 is no longer a disadvantage—it’s an opportunity. With today’s tools, the barrier to launching a startup is lower than ever. What still matters most hasn’t changed: solving real problems, earning customer trust, and proving people will pay for what you’ve built.
Or as the panelists reminded us: don’t overthink the tech. Just start building.
Full Transcript Below
Want to revisit a particular quote or share with a teammate? We’ve got you covered. Read the transcript here:
Jamie Gong
[ 00:03:11 ] Thank you so much. Good morning, everyone. Thank you so much for being here on Monday morning. I know the traffic's horrible. And yeah, it's getting cold here in Boston as well. But I'm very excited. This is my second year here at Startup Boston. And I'm a non-technical founder myself. So I'm very excited to be sharing this panel with my fellow founders. To discuss a little bit more, especially with all those exciting new AI tools. Everyone's talking about like vibe coding, how are we actually using those? Can we like replace our CEOs, all those AI tools? We're very excited to share more with this audience here. So we're going to start with around 45 minutes of casual discussion of things. We have a list of questions, but everyone here is a builder, so we'll just be talking about our experience building communities, building our product, and then we'll leave the last 15-ish minutes for questions. So if you have any questions that you have in mind, just raise your hand. So I would like to have my fellow panelists start with a little self-introduction about their background, how they have been building the startup ecosystem, and I guess how they have been using different AI tools to start this panel.
Jamie Gong
[ 00:04:11 ] Why don't we start with Sophia?
Sophia Lin
[ 00:04:13 ] Sure. Thanks, Jamie. And hi, everyone. I'm very excited to be here as well. Also a non-technical founder. Right now, I'm the partner of DATS Global, which we help founders and institutions to turn ambitious ideas into real ventures. I'm also the founder of AsquareLab, which we operate and invest in real estate adjacent sectors, business and niche tech. So myself, we built a bootstrap, a corporate solution, a corporate housing solution right now in around $15 million under asset without writing a line of code. And before that, I was in finance and got my MBA at MIT and then moved into entrepreneurship. Never wrote a line of code myself.
Sophia Lin
[ 00:05:04 ] I always use off-the-shelf tools and work with technical co-founders to build a product. With the current AI and a lot of automation tools, I think right now it is a really, really exciting time for those who are willing to take risk, ready to hustle, and have the tenacity to really start building. So I'm very excited to be here and discuss this with my panelists.
Karen Kelly
[ 00:05:31 ] Fantastic. Thanks so much, everybody, for coming. My name is Karen Kelly. I'm a non-technical founder. I have built businesses inside of tech, outside of tech, and I've worked in a lot of different industries. I've worked in advertising and healthcare, but I've been in tech for the last 10 years.
Karen Kelly
[ 00:05:48 ] And the reason why I am so excited about where we are today is that I have always sat in the seat of someone who cannot code. Whether I was the president of a company or the founder or CEO, it didn't matter. I lost control of timing, vision, and finances. So I am so excited this genie is out of the bottle. And I am running full force. Right now, I'm the co-founder of Launch by Lunch. We are a community for non-technical founders and fractional experts. We firmly believe that a rising tide lifts all boats. And so, if you want to talk about any of that stuff, I am your girl.
Elizabeth McWhorter
[ 00:06:21 ] Hello, my name is Elizabeth McCorder and I'm the co-founder and CEO of Thrive IEP where we take diverse data sets about students with learning differences and disabilities and we translate them into actionable, accessible plans for the adults that surround them and support them as well as for the students themselves.
Elizabeth McWhorter
[ 00:06:40 ] I used to introduce myself as a non-technical founder. I think I'd like to edit that. I would like to say that I'm a technical founder who does not code, because I think the nature of what it means to be a founder in this day and age, as Karen described, is completely different. The non-technical part of it has redefined itself. Our team is building our entire platform using AI coding tools, but that does not mean that we don't understand the technology that is organizing that information and that is allowing us to scale. We really access the deep technical knowledge, the information, the experience that we've developed over the last 20 years. 30 years each and deploy it very quickly and at low cost, without, as Karen described so nicely, running into the timeline and kind of the vision creep challenges that come along with a traditional engineering system.
Jamie Gong
[ 00:07:41 ] Awesome. So before we dive into questions, I actually like to do a little like power survey of the audience here. So how many people here consider yourself technical?
Jamie Gong
[ 00:07:51 ] Oh, okay. That's actually a pretty good representation. And how many people here is a founder or want to be a founder in the future?
Jamie Gong
[ 00:08:00 ] That's more people, right? We have more people who are interested in becoming, okay, how many people want to be a founder of a tech product? Let's put it this way. Okay, so we do have a great representation of people who are interested in building products, but don't necessarily have the engineering job of like coding. But I like to start the conversation with Elizabeth. so you mentioned that your team is building a product with non-tech like with with a lot of those new tools you just give us your full tax deck right now what are you building what is ai tools you're using how much do you guys buy code who's in charge of like writing the prompts any tips of like building a product from scratch using all those latest tools that we're seeing in the market right now Yeah, so we we started out using lovable as our primary space, I would say that was an amazing experience because.
Elizabeth McWhorter
[ 00:08:50 ] it was an enormous part of our learning curve because chatting with lovable actually taught me so much about what was happening within the product itself so that was part of like my shift from being a non-technical founder to a technical founder that ai codes We then kind of graduated, so we shifted into a Replit space, which is where we are now.
Elizabeth McWhorter
[ 00:09:14 ] We also use GitHub, we use HubSpot as our CRM, which also bridges into our product. But the vast majority of our work is happening in Replit. At this point, that's primarily led by our chief product and technology officer, which is Eric Falk, who is over there.
Elizabeth McWhorter
[ 00:09:35 ] And yeah, and almost everything is happening in that space.
Jamie Gong
[ 00:09:41 ] That's very exciting. And then I would like to ask Sophia a follow up because you mentioned a lot about like whether you use, you know, like actually custom software development versus off the shelf tools that help you build a platform. And it's really impressive that you guys have built like a gigantic portfolio without actually custom build any software. So can you just share a little bit more about your internal decision process in terms of build or buy, how do you decide whether to use some off-the-shelf material or use custom-built software when you're building your company?
Sophia Lin
[ 00:10:12 ] Thank you. Yeah, I think it's maybe good to define what you're building as a tech product or tech-enabled business. Well, I started with tech-enabled business, and then that is easy because your core. When you look into your core, it is about the business itself to serve the right customers. And you take all these off-the-shelf tools to help you streamline the operations, make it better, pricing better, and collect all the data to serve the customer better. So with the previous, through the past, I think, five to ten years, the off-the-shelf tools are already very good. But I think right now it's even more exciting because on top of our business right now, we're exploring to... create our own feature to our own customer bot, our own logistics bot, optimizing our operations. But because with all the past five, 10 years, the data we have accumulated and combined with the AI tools now we have.
Sophia Lin
[ 00:11:17 ] you are able to start experimenting this to build a customized one. But if you are just starting out without those tractions and data, I think off-the-shelf tools for tech-enabled business is good enough. You don't need to customize, build something that, you know, until you verify your product market fit.
Jamie Gong
[ 00:11:44 ] Yeah, I think that's an excellent point. Like, I think a lot of engineers I work with, they tend to build something and try to sell it. But sometimes it's actually better if you can test it with a product, like a lovable, like a replica. And then you actually sell the product before you actually build it out the whole thing. That's what we say. Sell the demo before you actually build it out the whole thing. A follow-up question to that is, how do you guys manage data? When you guys are using, you know, like third-party or off-the-shelf materials, because you mentioned that data is very important, especially with AI today. You have to have your own data that can help you train your bot better. So, how do you guys do the trade-off and how do you guys make sure they collect all those data, even though you're using some third-party software?
Sophia Lin
[ 00:12:22 ] that's a good question i have to say we haven't done a very good job on that so if anyone has a suggestion i would love to hear that but our business is not so data sensitive it doesn't work too much with personal data and when it comes to for example the logistics ai we're trying the logistics optimization tool we're trying to build it's really it's our own personal our proprietary data that we just collected over years so So that is easy for me, for us to do that. In the past, I've done FinTech Strata when it comes to data privacy was extremely important. And I think at that time, forming early partnership with data providers, with institutions is very important. but that requires a lot of effort. So when you go into, whether you have already decided which path you are going or you are still thinking about the venture you are building, I think incorporating those data privacy early on, like how, like the data barrier you will have to encounter is one of the.
Sophia Lin
[ 00:13:30 ] key things you want to start thinking before you really work on the product.
Jamie Gong
[ 00:13:35 ] Yeah, it's very interesting because I'm building in healthcare right now. And Karen, you mentioned that you worked in healthcare before. We all know the word HIPAA and how challenging it can be. And then as you're working with health systems, with SOC 2, HITRUST, and all those words, it just rings a bell in my mind. I'm like, oh my God, what are we going to do with our patient data? i think this is a great segue for us to dive in a little bit deeper into managing tech projects and tech teams so karen i want to get your perspective since you have like managed multiple different teams and different startups when do you think a non-technical founder like myself need a technical co-founder And how do you find those people and make sure you work well together when you don't necessarily speak the same language?
Karen Kelly
[ 00:14:16 ] I love this question. We talk about this a lot at Launch by Launch and with our community.
Karen Kelly
[ 00:14:23 ] So number one, you don't need a technical co-founder. That is my opinion. I'm sure people would disagree with me in this room, but today in 2025, to get a product out to market, to build an MVP, to put it in front of your audience and to learn. Whether or not the world is ready to pay for your product, you don't need a technical co-founder. You can build these things in a week on these products. I'm happy to show you how.
Karen Kelly
[ 00:14:46 ] So I would say, and we see. Great news. We see tons of evidence that there are VCs out there who are funding pairs of technical, non-technical co-founders. That's really exciting. Oftentimes it's two women, you know, maybe one's a product expert, maybe one's a marketing expert, but neither one of them has ever really written a line of code. So I would say that's a huge change. I mean, that's a huge change even from a year ago. You know, two years ago, I outsourced an entire product that I wanted to build. I made a lot of mistakes. But unfortunately, like, I couldn't make them myself. I had to pay, literally, pay for those mistakes. Um and like, I said, you know, you lose control of a lot of things when you do that. So, I think, you know, what we advise our members in terms of, well, how do you know when you need when you need further technical advice or you need a bit of low code? You know, what do you do? Or, you just have a question. Like, a lot of times with non-technical founders, now with the tools that are out there, you know, you can get really far and then you get to a decision point and you have to either like go left or right.
Karen Kelly
[ 00:15:50 ] You're like, 'I just need to know which way to go here' and they just need a little bit of guidance. Or, like, take two steps to the left and take one step forward, and like, 'Got it' and they run another 10 miles. Right? But then another question comes up. So, one of the things we've been talking you don't need a technical co-founder for that.
Karen Kelly
[ 00:16:04 ] You might need a fractional CTO, but I also think, you know, and I think eventually, all those organizations will need one if you have two non-technical. Eventually, you will. If you scale. But what we talk a lot about is finding really great advisors. And what we have found is Boston is a place full of people who have a lot of expertise, many of whom are in this room who want to give back and who want to help. And they're excited to just be part of this new wave of growth.
Karen Kelly
[ 00:16:35 ] You know, a lot of times I think if you find the right advisors, and again, that's part of what we aim to do is help make those connections, that's really important. And I also think that eliminates another problem which is finding a technical co-founder. I mean, I went through this for a long time looking for someone and it's very hard. It's not just, and this is really important. I think it's not just— do they have the skills you need? Or do they have experience building the type of product you want? It's what is their personality style? Is it complementary to yours? You know, are they more like— do they love details and spreadsheets? And you love marketing? Aside from— you know— just their core skill set? So much of it comes down to personality. I once worked with a CTO who just like really hated getting up out of bed before 9 a. m. But like, I'm ready to roll at 7:30. I'm like, let's go. Let's scrum. So, you know, that was that— like tough, too. And he would stay up late. And I'm like, I'm not available after 9 p.m. So there's all sorts of things to finding the right co-founder, which is why so oftentimes it is the source of companies breaking up.
Karen Kelly
[ 00:17:37 ] So I think it's great because it really widens the gate by which you can choose someone to build a company with.
Jamie Gong
[ 00:17:46 ] Yeah, I think this is fascinating because when I was in business school, we always talk about like how we are like all MBAs hanging out together. We never like meet the engineers. And then there's like always this joke that, oh, we should just like walk across the river to find engineers. But engineers are just like hanging out by themselves as well. They can't find the business people. But now, when I talk to MBA students today, they are increasingly using tools like to build their MVP], and then go find maybe a funding engineer or someone to build out a whole platform out for them after they already have the demo tested out with their real customers, which is really amazing. Cameron, you mentioned that you have a chief product officer and technology officer. Can you mention a little bit more about like, did you guys start working together from the beginning? What was the journey like? How did you move from lovable to actually having someone as a chief product officer?
Elizabeth McWhorter
[ 00:18:36 ] yeah so um we started working together almost a year ago um and did a lot of kind of experimentation so we did the the best like the recommended approach to exploring a co-founder relationship which is to work together and figure out whether or not you like each other whether your styles line up whether or not your like kind of like division of role makes sense And in this case, we worked together on contract for a period of time, and then shifted into having a conversation about co-founders. And that extended for, I don't. Eric, check me. This was like four months. We kind of had an ongoing conversation about what that would look like. How would we work together? What does being a co-founder mean? How do we define it? What does compensation look like? Um, and uh, and brought Eric on as our third co-founder about two months ago.
Elizabeth McWhorter
[ 00:19:38 ] Three months ago, at this point, and at that stage, we were um, past the point of the plans changing every two weeks.
Elizabeth McWhorter
[ 00:19:49 ] Which lasts for far longer than I expected it to. And we were at a place where our plans were fairly stable. Like we could look out maybe four weeks and say, 'This is what I think is going to happen.' I think we've just turned a corner where we can actually look out. Three months and we can project out a plan which, for the planner and me, I'm very excited about. But that also makes the the definition of the role of your technical team easier to actually look at. So earlier, you were talking about, what does it mean to manage a team? It's really hard to manage a team when everything is changing every two weeks. And it's really, really hard to project with.
Elizabeth McWhorter
[ 00:20:36 ] With enough security, what those deliverables need to be, um, how do you hold people accountable when the ground is shifting underneath you constantly? I think part of what I would advise is, in this day of AI coding and tools, where you really can build something, test it, and see if it's viable. To go as far as you possibly can without bringing on an engineering team, where they're expecting to be able to look at something and say, 'I'm going to build this over the next five months,' because it's changing so fast. It's not a good investment of time, money, or equity. To to try to aspire to that when you have other choices.
Jamie Gong
[ 00:21:19 ] That makes a lot of sense. I have a follow-up question to that: so you mentioned that you try to use all those like vibe coding tools yourself, and then you bring on this technical co-founder. How do you learn the language or the skills as you're like building out a company while trying to navigate all those, you know, ambiguity in the process, while also learning those tools to make sure that you can communicate with your chief product officer or like chief technology officer— as a like a founder who's like, you know, non-technical in the beginning, but transformed to be someone who's more technical in the process.
Elizabeth McWhorter
[ 00:21:49 ] I mean, that's the fun of it. Like if you're here to be a startup founder and you aren't ready to learn multiple new things simultaneously.
Elizabeth McWhorter
[ 00:22:00 ] While constantly having that ground shift under you and questioning what does it mean to be a CEO anyway, you're in the wrong ballgame. Early-stage startups is a constant state of learning new things.
Elizabeth McWhorter
[ 00:22:17 ] adjusting your expectations of yourself and the world around you and how you think things are supposed to work because they probably don't work that way in the context that you're existing within.
Elizabeth McWhorter
[ 00:22:26 ] And so one of the best parts about Loveable actually was chatting with it. like i use that like i don't know if it's still this way but like you can do free chats in perpetuity basically without getting charged any of your tokens and i would just ask it a million questions I'd be like, do you actually think this is the right thing to do? And it would be like, yeah, and this is why. And here's how this lines up with your SQL database. And I'd be like, oh, SQL database, cool. Oh, edge functions, cool. And I would just slowly look up the words and figure out how to speak the language. Because functionally, it's like learning a new language.
Elizabeth McWhorter
[ 00:23:06 ] As long as you can see how the puzzle pieces fit together.
Elizabeth McWhorter
[ 00:23:10 ] It's not that hard.
Elizabeth McWhorter
[ 00:23:14 ] And it's a lot of fun. And it's deeply, deeply empowering to be able to actually see what's happening and how it evolves and solve the problems on your own. Especially, I would say, as a woman founder and a first-time woman founder, when you are so often expected to not be able to do this by yourself and that's enormous.
Jamie Gong
[ 00:23:41 ] Yeah, I was reading some newsletter, I think, and I think there are people like just building business off Louisville, just like purely vibe coding on Louisville, and they create literally like millions of dollars in like annual revenue. It's really liberating and just empowering to see how those tools can be capable of producing. But it's also very interesting because I think we our company uses V0 a lot for prototyping of like our UI/ UX interface, and it's very interesting that how that sometimes just really a V0 meaning that it's very easy to start from scratch, but it's actually much harder if you want to make it like scalable. So this is a question for Karen. How do you think about managing those technical debt and scalability issue as you are building a company that hopefully we all want our company to scale to thousands of millions of users? how do we do this trade-off between you know or have something that's like really easy and vibe coded and like just really easy to produce versus like actually thinking about the architecture the cloud infrastructure and all those things to make sure that you're building a scalable software Well, first,
Jamie Gong
[ 00:24:42 ] I think you're bringing up a really good point about how easy it is to vibe code a lot of these things.
Karen Kelly
[ 00:24:46 ] And one of my first thoughts after I, you know, actually, if I take a step further back, I think one of the biggest things that holds people back to what Elizabeth was talking about in terms of, like, actually getting into these tools and your question about, like, is just a psychological barrier of like, this feels really overwhelming. And like, I'm told to use 12 different tools. Like, which tool should I use? And I only have 25 minutes. And which one should I pick up? And I think that is holding a lot of people back.
Karen Kelly
[ 00:25:14 ] Um, unfortunately, um, I wish it didn't but, um, you know, that's that's the nature of AI right now, um, which I also think is really important to, like, find somebody else who wants to learn something, pair up with them, buddy up with them, and just learn one thing together. Elizabeth and I both sort of turned on to these tools at the same time. So it's been fun to kind of go on the road with her, and, in the background, we're like, 'What does this mean? What are you doing? I don't know what's happening.' So it's really fun to, again, that's why we're also focused on community at my company because I do think it's much easier.
Karen Kelly
[ 00:25:47 ] Getting into more around this idea that of scaling first— if you can vibe code something in a day or a week, someone else can too. And I think that's a really important part of this discussion for all of us to have and to think about. Because ideas are cheap.
Karen Kelly
[ 00:26:06 ] Really, ideas are worthless. Vibe coding an idea in a week is a little bit of something. But to be honest with you, you have to be all the same principles apply to a vibe coded business versus a regular startup or any other small business in the world. You have to be solving a problem that people are willing to pay for, and they have to like your product. They have to like using it. And so I think those things are really important to keep in mind in terms of when we think about going out into the world, okay? If you can do it, somebody else can too.
Karen Kelly
[ 00:26:37 ] But then, I think in terms of scaling and things like that, there's a lot of discussion out there. Well, in order to scale, you're really gonna need to use some of the more legacy products. You're going to need to bring on a technical someone.
Karen Kelly
[ 00:26:50 ] To be honest with you, So I'll use a word— Supabase. A lot of you are probably cringing or you're like, 'I don't know what that is.' It sounds terrible. It's just a database that connects through a lovely and it just holds all your data. So if you want to upload profile pictures or if you want to upload videos, like, you just connect it. It takes literally a minute and a half. It's super easy. And then you can do all these really amazing scalable things. I mean, it's a cloud-based database. So you can go really, really, really far just by making that one connection.
Karen Kelly
[ 00:27:21 ] So again, I see more opportunity. Yes, at some point. If millions and millions of people are using your product and you are making hundreds of millions of dollars, yes, absolutely, you need to be investing. But at that point, you will have investment dollars or you will have revenue, hopefully revenue, revenue over investment dollars, and you will be able to find to hire people who are truly experts and can help you get to that next level.
Elizabeth McWhorter
[ 00:27:45 ] I just add one thing: I think scale is a great problem to have, but when you're just starting out, don't worry about scale. Don't worry about building something, testing something, or building a new version of that thing. If you're thinking about scale and you're very early, you're thinking about the wrong problem. You have to be thinking about whether or not you've got even something that's close to product market fit, and whether or not you can actually do the reps to test that with other people. So don't worry about scale at the beginning. Worry about building something that people want. Then worry about scale later.
Elizabeth McWhorter
[ 00:28:25 ] Don't start with scale, though. Totally agree. Which is hard, because we're supposed to think about scale all the time, but don't.
Karen Kelly
[ 00:28:32 ] One other thing I'll mention, too, some founders have brought to us is: Well, I have it as a web app, but I want to move it into the app store. And it's like, it is such a pain in the neck to move apps into the app store. And it's money, and it's a pain in the neck, and it's something to maintain. And it's like, don't do that either. Really, build a web app. Get people to just save the link on their phone or engage with it on their desktop. Do not worry about these things. Just start super basic. What is that one core problem? Build it, bring it to the world, and see what it says.
Jamie Gong
[ 00:29:04 ] That's an amazing, amazing point, actually, because that's something me and my CTO actually discuss a lot. It's like, oh, do we need an app? Like, why do we need an app? Do we want to go through the whole process of the app store approval process? Do we want to actually invest? time money and effort into like building an app or can we just do like a more user-friendly interface because honestly a lot of things with ai today you can do a lot of things with texts with phone calls like very like very manually but all those like ai call agents are just like really amazing and we automated our interview with AI today and that just improved our interview completion rate by like 50% just because we can do things async without actually human in the loop. So I think there are a lot of amazing things that we can do with AI today that's just like unimaginable a year ago when I was building the prototype of my company which is literally Notion pages and Google Forms. And that was just literally a year ago. Yeah. So I have a last question on managing the tech team and projects. And that's for Sophia. I'm curious for the business that you've been in, like in FinTech, in real estate investment. What are some red flags that people should watch out for if they are looking to work with someone who's technical or working with a contractor, sometimes because they probably don't have the internal capability of hiring an engineer or they're not necessarily building a software business.
Jamie Gong
[ 00:30:17 ] It's a tech-enabled business.
Sophia Lin
[ 00:30:19 ] yeah um i think we've all heard the phrase the higher fast and fire fast um i think one red flag that i've constantly watched out and i've constantly proved i find it proving correct is that when you whether you hire a contractor or you work with with your technical co-founder or some vendors, if they cannot deliver when they say they can, that is something that you want to start thinking. I think a good vendor, technical co-founder, contractor, they should have a good grasp of the resources and timeline. They can accomplish a feature. And if they constantly miss the mark, it's really hard for you as a business side of founder to communicate with your customers, with your investors. And you keep missing the timeline.
Sophia Lin
[ 00:31:19 ] People lose confidence in you. You become uninvestable. you become like customers will walk away because they cannot trust how reliable your product is so your co-founder your vendor your contractor partners how reliable they are how good they can give you a good estimate is very very key when you are building an early on product And don't get me wrong, 90% of the time the tech team will always ship a feature that needs to be iterated or they might miss the deadline by two weeks. that's totally normal but if they are like giving you a time and they miss a total two three months i think that's the time you need to have a conversation and if it keeps happening they don't have a good knowledge of how the the time estimate and resource estimate and that's will that will make your
Sophia Lin
[ 00:32:21 ] business very very difficult early on wow two to three months has this ever happened to you guys yes And this is why we vibe code.
Karen Kelly
[ 00:32:34 ] Exactly right.
Karen Kelly
[ 00:32:35 ] It's that control piece, right? I mean, again, we're going to probably be late too, but at least I know why I'm late, you know?
Sophia Lin
[ 00:32:44 ] I want to add, I think here I'm not saying like a big feature shift. I think I'm talking about small, like, okay, maybe take it back. I think a good partner founder, they also, they should be also very good at communicating the steps, the roadmap. and it's not like oh we we have an end goal and then we have this big feature but how do you break down like what timeline they're they're going to hit because you will always you always have to go back and review the product feature and how this works within your business you might be changing along the way it's not like you ship a big feature and to the market the customers don't like it you want to test it out small pieces piece by piece and i think breaking it down is helpful for both business side and technical side Yeah, I think that's really interesting because Elizabeth kind of alluded to it, is that when your business goal is kind of like moving every two weeks, it's actually really hard to have like a three
Jamie Gong
[ 00:33:43 ]-year technical roadmap of like, oh, Those are all the features we're going to ship within the next three years. It's incredibly hard to plan for that when you don't have like something set— then, oh, this is the customers we're building. This is the contract we have, and then now we're actually going to be investing, and then building building the product— very very interesting. Um, okay. Now I want to move on a little bit into talking about, you know, like what the things that founders should be cautious about when using AI. We talked about all those amazing things using AI, and all those like new tools people can be using to ship something really easily. So Elizabeth, I'm curious what AI and automation tools that you see, uh, founders should be cautious about adopting when they're like experimenting and like coding today.
Elizabeth McWhorter
[ 00:34:24 ] I think, in general, what I would say across the board with these AI tools is: assume that they're going to lie to you because they will. They'll say they did something and they didn't.
Elizabeth McWhorter
[ 00:34:43 ] Maybe not trust, but verify. But like, uh, sort of trust and verify deeply. Um, I think and building that assumption in that they're going to say that they did something that that didn't happen or isn't happening in the way that you think it's going to.
Elizabeth McWhorter
[ 00:35:01 ] I think the other kind of mindset, and I know I'm not quite answering your question, but I think the tools are changing so fast that every point in time answer to this question is going to shift.
Elizabeth McWhorter
[ 00:35:14 ] Two to four weeks later.
Elizabeth McWhorter
[ 00:35:17 ] But assume that you're giving instructions to a very junior staff member. Be explicit, ask lots of follow-up questions, clarify what you're looking for, and be okay with going back and having them undo work the same way that you would with somebody who's just out of college, who's joining your team, who's never been in a formal work environment. So I think, as long as you assume those things going in, you'll catch the mistakes and you'll be able to choose which tools actually work best for you. Because I think that's the other beauty of this time is that there's so many tools out there that you can experiment with them. Use all of the free trials. Give them all a try and see what you like and then just land on that one and stop trying other things until it stops working for you and then find something else.
Jamie Gong
[ 00:36:15 ] Yeah, that's an excellent point. I still remember when I used Cursor for the first time, and then Cursor would say, 'Oh, I did all those changes for you,' but when you actually go into the code base, none of those are done. July.
Elizabeth McWhorter
[ 00:36:25 ] Like, none of it is in there.
Karen Kelly
[ 00:36:28 ] Yes, I was just going to say it does want to please us. It wants to please you. And it's not trying to lie to you. It thinks that it did it, but it didn't. But one thing I will say in general is that you. AI is very good at fixing itself. So, but you have to ask it to. So, if like, I've untangled a lot of spaghetti-vibe coded products by simply asking the AI, like, 'I don't see this thing that you said that you put here. Can you please tell me why it's not there?' And then it'll generally come back and say, 'Oh, I see what you're saying. You're right. It's not there. Let me go look for the reason why.' And more times than not, it can figure it out itself. So I would say: But we are, as humans, interacting with technology for 20, 30, 40 years, we're not taught to ask the technology to figure out why it's broken. We don't ask our vacuum cleaner why it won't start. But you can do this with AI, which is really, really cool. So it may lie to you, but it also can fix itself upside.
Elizabeth McWhorter
[ 00:37:27 ] You can also ask other AI tools to validate what the AI, what the dysfunctional for the moment AI tool is telling you that it did or what the next step is. So you can ask the tools to critique each other, which gives you better answers. short-term and the long
Karen Kelly
[ 00:37:44 ]-term like a chat GPT hey I'm having this problem with lovable I can't fix it can you please tell me what some of the common reasons are why lovable why this happens in lovable it'll spit something out and then it's like okay great can you can you give me a prompt to give to lovable to untangle it sure here you go sweep that over to your vibe coded tool. So I do a lot of that. I think we all do a lot of that. That's a great hack when you get stuck.
Jamie Gong
[ 00:38:07 ] Yeah, one hack from me is that it's not even with my coding tool. Sometimes I use Figma for design or use like Framer for like website. I would like just screenshot things that I can't figure out and feed it into ChessGPT. Like, oh, this is what I'm trying to do. Can you tell me what exact steps that I need to take to make this like chart, whatever. And then ChatGPT, oh, like, go there and click on this button, or go there and then, like, right-click on this thing, and then figure it out for me. So that's, like, really easy. I don't need to go into, like, the pages of, you know, like, a user's manuals or instructions. And then ChatGPT can just, like, digest it and give me the answer, which has been extremely helpful. I think, Cameron, you kind of alluded to this a little bit about there are problems that AI and technology can solve, but there are problems and challenges for a startup company that can't be solved by AI. So I would love to hear your opinion on as startup founders or aspiring founders, what should we focus on as we're building our business, especially for early stage startups? And what are some issues that we can't just like wipe coat away and then we have to really, really figure out as we just get started?
Karen Kelly
[ 00:39:07 ] I think anytime there's a human, you know, you're delivering a human-based service, like you have to be really, really careful.
Karen Kelly
[ 00:39:15 ] Most people are not using chat, even ChatGPT, which we all kind of think of as the entry-level AI. Most people, like less than 30% of America, has ever tried it. So I think we're in this bubble here in this room and in the tech startup scene. We kind of just assume that everybody's where we are. Most of the world is not. I have had like great, like, I wouldn't say joy, but like laughs over, like, I feel like I live in two different worlds. I live in the tech world, but then I'm also a mom in the suburbs. So like, I go on and I watch what parents are saying about AI in ChatGPT. I watch a company, which I will not name, but it's a big company that teaches AI, try and target market moms in the suburbs, and it's like. You only know ChatGPT. Meanwhile, teenagers are starting businesses with like this name of this company. Like tomorrow morning, we're going to get started. And the comments, the moms, that's so great for the teenagers. I don't care about AI. Like, why are you showing this to me?
Karen Kelly
[ 00:40:16 ] And it's just like. You totally missed the mark. So I think like that kind of clash, beware of the fact that number one, most people don't trust AI. So if you put something in front of them that's blatantly AI, most consumers. Are not gonna— i'm making a blanket statement but many most consumers are not average Americans are not going to want to interact with it. Um, so I would say be really wary of that. I would also say you might want to veer away from saying the word AI. I know it's one thing to talk to investors and they might want to hear that. But when you're talking to customers, you actually probably want to veer away from it because there's so much negative connotation. Everyone still thinks of the Terminator. It's really like there's a lot of bad vibes about AI out there in the general population. So I would be really aware of that. And I would start by building a product that's made better by AI instead of saying, 'I have built in AI and here it is to interface with consumers.' Power your product with AI, but don't make it the, don't make it the any kind of real interaction with the humans using it yet.
Karen Kelly
[ 00:41:20 ] We're just, we're really not there.
Jamie Gong
[ 00:41:22 ] Oh my God, 100% agree. So I work in senior care. So I work with a lot of caregivers. And I mentioned that we use AI to automate our interview process because a lot of caregivers are always on the go. They're doing like 12-hour shifts. So it's incredibly hard to schedule interview with them. Because we can't do interviews at like 2 a.m. m you'll be surprised by the time of people who call in and do interview with our ai but we also have like a human in the services right like i i'm like picking up a customer service call myself so there's one day actually last friday i got a call from like one of our caregiver and she's like so are you a robot And I just couldn't, because we do have AI interview tool, which is a voice agent. And we say, oh, I'm Ann. I'm your AI interviewer. And then we are talking to you. But we also have humans who are actually talking to our users and our customers who are the caregivers. So it's incredibly hard to balance because people sometimes have this fear. They're like, oh, I'm talking to this robot. I don't feel like this company's personal enough. But you also want to improve the efficiency and to make sure that you're meeting people where they are with all those new tools that we can have. I think, Karen, you mentioned a little bit— I'm also interested in Elizabeth's perspective on this, especially as you mentioned that as a non-technical founder, a woman founder, building a technical product, and how do you think about in today's fundraising environment, how do you build credibility
Jamie Gong
[ 00:42:35 ] with investors? How should we be thinking about fundraising? I love what you said about revenue being the best way of funding a company. absolutely believe that but as we are building our company a lot of us need to raise money a lot of us need to talk to investors so elizabeth from your perspective how can what are some like strategies and tactics tech non
Elizabeth McWhorter
[ 00:42:52 ]-technical funders can use to talk to investors and build credibility I mean, I think understanding the pain point, understanding, demonstrating credibility in other ways. So, like, we know our pain point. Like, we have done our due diligence. We understand who our customers are and what their needs are, and we're able to bring that to the table. And then we're also able to show that we built stuff. So like if I think it's I think it's different when you haven't built something.
Elizabeth McWhorter
[ 00:43:28 ] There's an angel investor, early seed investor who's always like, show me what you built, build something and then come back to me. And now you can build something in a week. So the conversation is starting to change. I think the, and then the other piece is like you're inviting somebody into your company.
Elizabeth McWhorter
[ 00:43:49 ] you're inviting them into an opportunity, and you're inviting them into a team that you believe in. And so if you don't believe that your team is the right team, they're going to feel that. and they're not going to be interested in investing because they're going to know that you don't have confidence in your team and so, but I have 150 confidence in my team and I bring that into every conversation because I know that we can do this. And so that credibility, and I think this is the reframe of like, 'Oh, you don't have a CTO? Like, what's wrong with you?'
Elizabeth McWhorter
[ 00:44:30 ] I like it— it's a it's shifting into a completely different conversation, and I think we have the opportunity to start to really push the envelope as founders because investors need founders to.
Elizabeth McWhorter
[ 00:44:46 ] Like, this is a two-way street.
Jamie Gong
[ 00:44:50 ] I love that. When I was like fundraising, I always think about how I'm also giving the investor an opportunity to invest in an amazing company. I mean, when there is like the power dynamic and also like what's the challenge of like talking to all those like super big name funds, sometimes you feel like vulnerable as a founder. But you have to believe that you are actually inviting them into the company by giving them a great opportunity to work with you. And the confidence is really something that's really matter in the process.
Karen Kelly
[ 00:45:16 ] I would add one thing there, which is that if you are non-technical and you don't have anyone technical on your team and you're sitting down with investors, number one, they probably know that. So that's a good sign that they've taken the meeting. But number two, you want to head it off at the pass. Learn what you need to learn. Know your tech stack, understand how your product works, and lead with that. When you lead with that, you've got the t-shirt on, right? You've got the credibility and the credentials that you need. Use the language. Learn the language. There's work to be done there. But when you lead with that, I think it just... It changes how they are seeing you.
Jamie Gong
[ 00:45:51 ] That's an amazing wrap to our discussion here. And I would like to open it to questions. So wow, I see many hands. OK, we'll start from the one in the back.
SPEAKER_5
[ 00:46:01 ] Please make sure you have a microphone. There's two of us walking around with them.
SPEAKER_2
[ 00:46:07 ] Hi guys, thank you so much for the discussion. I have three questions, if that's okay.
SPEAKER_2
[ 00:46:14 ] Sorry.
SPEAKER_2
[ 00:46:16 ] It'll be fast. First one is, since the barrier of entry is lower now thanks to vibe coding, what's the type of traction that these vibe coding founders are showing to get funded? The second question is that it was mentioned that ideas are cheap. Some of these vibe coded businesses are starting as a passion project. I have one.
SPEAKER_2
[ 00:46:39 ] At what point should these passion projects require NDAs when having discussions? And the third one is for you, Karen. Hi.
SPEAKER_2
[ 00:46:50 ] I applied your suggestion during the weekend of getting users. I got them at a baby shower. The feedback was overwhelming that they want the native app. They don't want a web app. So at this point... Should I continue getting feedback or should I start going through that grueling process of a native app?
Jamie Gong
[ 00:47:09 ] Okay, who wants to take the first one on funding?
Elizabeth McWhorter
[ 00:47:14 ] So in terms of how do you position traction?
Jamie Gong
[ 00:47:16 ] Yeah, and then what's the threshold of getting funded?
Elizabeth McWhorter
[ 00:47:22 ] I don't know. I feel like this is the question that everybody wrestles with all the time. I would say customers, preferably paying ones.
Elizabeth McWhorter
[ 00:47:31 ] is your most significant traction moment. I think it's part of why pushing for that first paid customer is such a big deal.
Elizabeth McWhorter
[ 00:47:42 ] And then I think the other traction numbers really depend on your product and what does customer success look like for you. But I think the biggest one, if you can focus on anything, it's paid customers. Because people will use something for free. Do they convert to paid?
Karen Kelly
[ 00:48:02 ] That's the answer to question number three. Go get those people to pay you money for their product. There's nothing inherent about that product that needs to be on a mobile device. So either they love it. If they love it, they'll open up their Chrome browser or they'll open up their desktop. Bottom line, any customer. If they love what you're delivering, they will go out of their way to use it and they will pay you money.
Karen Kelly
[ 00:48:23 ] That's the bar.
Jamie Gong
[ 00:48:24 ] Sophia, do you want to take the question on NDA?
Sophia Lin
[ 00:48:27 ] I actually want to add on to the first question. I haven't done fundraising for a while, so I don't know what the market right now is, but we do.
Sophia Lin
[ 00:48:36 ] periodically go out and we like scout and like invest in companies or buy companies. What we are looking for really low threshold— like if you have 20, 10, 20 paying customers and making some good money, like some product market fit, that we think we can include in our portfolio and have a good synergy with what our existing ventures. We're interested. um and by that i i i think an interesting landscape ship right now is that uh it is true idea is before idea worth nothing because you have to go on and build. Right now, I think idea is still worth nothing because you have to go out and get your customer to pay. Customer paying, paying customers is a very, very powerful market validation because everyone can build a product right now. but you probably are the only person or one of the few people that understand the target market you are addressing very, very deeply.
Sophia Lin
[ 00:49:44 ] And that is the insights people like us, we don't know. Like I wouldn't be able to understand all sorts of market, but I would be very interested in what you discover and what kind of product, what kind of problem you are addressing to get your customer to pay. And to me, that's a market opportunity.
Elizabeth McWhorter
[ 00:50:04 ] the other question he had about the nda i have a perspective on i have a very strong perspective on this well then why don't you go first i do too it might be the same one like unless you're applying for a patent yeah never never an nda no absolutely not If you're that worried about somebody stealing your idea, your idea probably isn't good enough.
Karen Kelly
[ 00:50:25 ] Or it's so early. It takes so much. There's so much friction between an idea and success. So many layers.
Elizabeth McWhorter
[ 00:50:32 ] And it changes the whole conversation. I've signed two NDAs since founding the startup, and they were both for companies that had deep tech.
Elizabeth McWhorter
[ 00:50:42 ] patent application hadn't been submitted yet and those were the only times that I signed NDAs and felt like it was worth it everything else was like yeah like why are you doing this like it was a flag for me let's take the question from the my left hand side Thank you.
SPEAKER_7
[ 00:51:06 ] Hi there, thanks so much for your comments. It's also really awesome to see an all-women panel. My name's Claire. I do fractional research and product strategy for startups. And I mostly work with women-owned and minority-owned startups. And I work with a number of founders that have disabilities. And often when they hear you can do this in a week what they hear is this isn't built for me and so i'm curious if you have any thoughts on accessibility and some of the tools you're talking about um and like where we can be lowering the barriers of access for founders that maybe have visual disabilities or cognitive disabilities or neurodiversions Yeah.
Elizabeth McWhorter
[ 00:51:44 ] So I think it depends on the disability. So, as somebody who has ADHD, this is a very challenging and amazing time to exist. I'm highly distractible, and so it means that, like all of these tools and this ability to get these very fast repeated dopamine hits, are both amazing and completely overwhelming. I think there's like a cognitive and a planning barrier there. I think a lot of the language-based platforms can be really challenging when you're talking about individuals with dyslexia, and when you're talking about ADHD as well. There's a lot of ways that we could be better designing AI coding tools and other kinds of low-code or kind of faster tools. I think the other piece is thinking about how you actually break down the tasks.
Elizabeth McWhorter
[ 00:52:47 ] So, like when when you're when you're building something and it's moving so fast, and you don't like to move that fast when you like to have all the details worked out at the beginning, this can be a very scary time.
Elizabeth McWhorter
[ 00:53:04 ] So I think there's a number of issues that we have in terms of that— this isn't built for me. I do think we have a lot of opportunities in terms of where disability tech is going. to see how this can actually be an accelerant to accessibility and so I think there's there's a whole group like a whole group of dedicated specialized funders as well as accelerators and things like that that are very much focused on disability tech that I'm really excited about and ways to to point out how current technology is actually not accessible as well so so I think there's there's opportunities there but I also think like to go back to Karen's point from earlier like a lot of this is about community and so the same way women are also often not viewed as being kind of ideal founders for investors that disabled individuals face similar barriers and so when you build community and you have people who are coming together who are trying to figure out how to navigate a new space and a new set of tools, doing that with other people who kind of face similar challenges can be deeply empowering.
Jamie Gong
[ 00:54:21 ] Front row here.
SPEAKER_8
[ 00:54:29 ] Hi, everyone. My name is Nicholas. I am a founder of a fashion tech startup.
SPEAKER_8
[ 00:54:35 ] It's an AI company, and I just wanted to ask her. She mentioned that people don't yet trust AI. And we're trying to position our company. We're creating fashion agents— agents for shopping, for fashion and apparel. So, how is the position like, considering what you're saying, the people you don't yet trust?
Karen Kelly
[ 00:55:01 ] That's a great question. So I think one of the unique things about what you're doing is that what your company does isn't like it's going to go—assuming you prevent it from actually buying the clothes without them approving it, right? Because I guess what you're saying is that these agents go out and they have intelligence about the individual and they can kind of pick out clothes. Is that basically what you're saying? Yeah, okay, okay. So it's customized so so that kind of intelligence sounds really sweet. I think I would not build it in such a way to start off that like we're gonna go find the clothes. We already know you like these brands. We already know your size, and we're going to buy it for you. Of course, that'll scare people. But I think if you send agents out to go find styles based on XYZ for you, customize, that's awesome. I want AI to do that for me. And then bring it back to me, and then I'm going to make some selections. And then when I'm good and ready, maybe I will let it go make a purchase on my behalf. i think it's okay to do it that way but i don't think you need to i wouldn't lead with ai in your marketing or in your company name or anything like that we don't care i don't care that it's ai i just know the benefit and i think that's really important what is the benefit of your ai for any company
Karen Kelly
[ 00:56:05 ] put that out front saving time whatever it is put that out front This gentleman has had his hand up for a while. Yeah.
Karen Kelly
[ 00:56:16 ] No, that's you. Wait for the mic.
SPEAKER_6
[ 00:56:22 ] Hi, good morning, thank you. I have two questions. First, Elizabeth, you mentioned that you graduated from Lovable to Replit. I was wondering what Replit can give you that Lovable perhaps was falling short. And then my second question was for Karen.
SPEAKER_6
[ 00:56:38 ] You mentioned also that ideas are cheap and right now if you can build it in a week or a day, somebody else can. How do we defend our new company if that's the reality?
Karen Kelly
[ 00:56:51 ] Do you want me to go first? Sure.
Karen Kelly
[ 00:56:53 ] You defend your company with paying customers. That's the bottom line. The beautiful thing, too, is that not only is it fast to build, it's so cheap. Typically, that's been a big barrier for founders who are non-technical is the money it costs to outsource something like that. You don't have to outsource it now. Let's say you spend... 200 a month on tokens and whatnot to build this thing maybe a little bit more um you know that's not really a barrier for most first world people anymore so um so yeah so get it out into the world and then like get get people you know that that's it's it's never been easier or more important to have like active customers who are paying you money All about the customers.
Elizabeth McWhorter
[ 00:57:34 ] Always. So we switched from Lovable to Replit in large part because of the database integration. So Supabase is great. Um, but it didn't allow us to reliably uh interact with the large language models that we wanted to be using, so we kept on running into challenges there. Therefore, Replit was just a better fit for us at that point in time.