Listen to the episode here (30 mins)

We recently had a 30 mins podcast interview with Richard Bradbury on BFM 89.9 - The Business Station. We've been a big fan of the show for ages and were super honoured to be on. BFM is a leading business radio station based in Malaysia.

Here's a couple of the highlights of the conversation
- Stories on our early days building in the AI space
- How we’re thinking about global competition
- What Antler’s accelerator is really like
- Running business experiments in Cleve

If you prefer listening to the episode, we've got the Spotify and Apple Podcasts links below, or if you're more of a reader there's a summarised article version too.

Listen on Spotify & Apple Podcasts now.

Transcript

Intro & About Cleve

Richard Bradbury: This is a podcast from BFM 89.9, The Business Station. Good morning! It’s me, Rich Bradbury, and welcome to Open for Business. By now, you should know it’s the show that dives deep into the journeys and ventures of entrepreneurs and business leaders. This morning on the show, I’m speaking with Ashvin Praveen. He is the co-founder and CEO of Cleve.ai, a local startup that’s aiming to change the personal branding landscape with their AI-powered content creation workspace. We’ll get into all of those details a little bit later on. He co-founded the business back in December of 2023 with his partners, who are here today, Lizzie and Rashad. With their combined experience in AI, product, and personal branding, they’ve built Cleve by combining the elements of note-taking with modern-day content generation to help creators scale their content efforts with the help of generative AI.

After just six months since their inception, their platform has now grown to have over 120 weekly active users and is steadily growing. They also secured funding from Antler, the global early-stage VC firm, earlier this year. Again, we’ll talk about that a little bit later on. With that, Ashvin is here with me in the studio, and we’ll be discussing the story behind Cleve’s inception, as well as the foreseeable challenges and opportunities ahead, considering this very competitive yet nascent industry. If you have any thoughts, you can reach us on our YouMobile WhatsApp number at 018 789 8899, or you can reach us on X at BFM Radio.

Ash, welcome to the show.

Ashvin Praveen: Hey, thanks for having me. It’s a pleasure.

Richard: How are you this morning?

Ashvin: Good. Yeah, I’m doing good now.

Why the name Cleve.ai?

Richard: Talk to me first and foremost about the name Cleve.ai. Where did that come from?

Ashvin: So, would you like the politically correct answer?

Richard: No, absolutely not.

Ashvin: Let’s be real. We were looking for a one-syllable, five-letter word that ends with the .ai domain. We shortlisted, like, 20 names, and Cleve was the most interesting one.

Richard: So, the three of you were sat in a room, frantically typing out names of five-letter words?

Ashvin: Pretty much. We used ChatGPT and stuff, basically asking, "Hey, give me a five-letter word, ideally related to personal branding." But yeah, that was a fun experience. That’s how you should start things up.

Where did the initial inspiration for Cleve come from?

Richard: Walk us through the initial inspiration and motivation behind founding Cleve. What do you think was the specific gap in the personal branding market that you were aiming to fill when you first started?

Ashvin: Yeah, so we’ve been working with personal brands for the last four years, actually. During university, we started a freelance business where we were helping business leaders and CEOs position themselves and build their brands, and also creators, like YouTubers. We saw this whole market growing—the need for personal branding and the opportunities that come with building a strong personal brand became more and more mainstream over the last couple of years. So, when we joined this pre-accelerator called Antler in October last year, we basically saw an opportunity to take what we were doing and our knowledge from the agency model where we were building personal brands and try to transfer as much of that as possible into the domain of AI.

Richard: When you say "we," you’re talking about the three of you already?

Ashvin: Oh, not yet. At the time, it was just me and Lizzie. The two of us started that business about four years ago, and Rashad joined earlier this year.

Richard: The business is obviously run around AI and generative content. People listening at home might be thinking, "Can’t I just use ChatGPT for that? Or can’t I just use Bard, or Gemini, or whatever?" What is it that makes you guys stand out from the rest of the crowd?

Ashvin: As we were interviewing creators and personal brands, we asked them, "Hey, do you currently use ChatGPT, Bard, etc., for your personal brands?" We realized that the big gap all these existing tools aren’t really filling is that they lack context—they don’t know who you are, they don’t know your tone, they don’t know your personality. As a result, you end up with a lot of content that’s very generic and fluffy. Nobody really wants to post that on their personal brand LinkedIn, for example. That’s where we saw an opportunity: if we could build an AI that has a lot more context, that’s connected to your social media profiles, that’s able to analyze all your past content, then we might be able to carve a niche in this market. We’re building personalized AI that really understands who you are, how you talk—we could make an AI Richard! And hopefully, that would be way more helpful to people who are building their personal brands.

Richard: So, to simplify your plug, one of the things that you do is plug into that existing person’s socials, read through their stuff, and get a feel for who they are and what they might say.

Ashvin: Yeah.

Managing privacy concerns and data security

Richard: Okay. So it’s a buy-in from the person who’s already subscribing to your product. One of the things I’m going to jump straight to here, which I might get to a little bit later, is privacy concerns, right? If you’re already signing up for something like this, you’ve agreed with Cleve to read and pull through my data, right? Is there a worry for some? We know that data leaks happen here in Malaysia. How do you protect the data that people are feeding into your LLMs or whatever it is that you’re using?

Ashvin: Fundamentally, with the data concerns, a lot of the social media data we’re collecting from users is public data. People just click one button to connect their LinkedIn. It’s all public data that anyone could get on their public LinkedIn profiles and stuff like that. But yeah, we do see data protection as a main priority. A lot of our systems are designed with scalability and data protection in mind first, before we do anything else. That was our top priority when we were initially setting up the product.

Securing startup funding without a product

Richard: Okay, that’s nice. Now that we’ve got some of the stressful things out of the way, let’s move on to another stressful thing: funding. You spoke about Antler just a few moments ago. We know that securing funding is often a critical step for startups. How did you go about raising the initial capital to start Cleve? What were some of the key challenges you faced during the fundraising process?

Ashvin: In our case, I’m quite grateful for Antler because it was quite a hand-held process. We joined their accelerator in October last year. It was 10 weeks of, honestly, the most intense 10 weeks of my life. We were going through multiple products, iterating, finding validation, and stuff like that. Basically, at the end of the program, we pitched to the Antler investment committee. It was that one pitch where we showed our validation, traction, business model, etc., and they were like, "Hey, yeah, let’s fund you." So it was 110K USD, the standard Antler check, and that’s basically where we got our start. At that point, when we were pitching, it was just an idea. Before joining the Antler program, we didn’t even have much of an idea. It was during the program that we iterated, figured things out, and got early validation.

Richard: So you had no MVP at that point?

Ashvin: No, not at all. We did put together some proof of concepts with ChatGPT, custom GPTs, and stuff. It was very scrappy, but it was good enough.

The Antler 10 week accelerator

Richard: I can imagine how that would have been stressful. Tell me a little bit about those 10 weeks. You said it was perhaps the most stressful period of your life. What was so stressful about it? Was it the fact that you’d never done something like that before? Was it the potential constant scrutiny? Was it the hours you were putting in? Tell me about it.

Ashvin: I think it was a combination. The biggest thing was that I’d never built a tech product before, and I’d never explored the startup space or tried to figure out how to build in the tech space. That was my first introduction, and the biggest challenge was really rewiring how my brain works because it’s very different from traditional business where you come up with a business idea, find customers, and sell. With Antler, it’s about validating first—it’s that lean startup kind of mindset where everything’s just a series of experiments. Your first business idea is probably not going to be the right one. Whereas traditionally, I would think, "Okay, find the best business idea, and then just go all-in." So, it was a lot of that kind of rewiring. We would be talking to customers in whatever ways possible, trying to find validation, testing different ideas, and then killing ideas really quickly. We went through about 30 ideas during the program, where a lot of them were killed at some point.

Richard: At some point, you’re going, "I don’t think my brain can create any more things."

Ashvin: Exactly. At the end of it, we found the personal branding niche, went deeper into that particular process, got more validation, and that was great.

Richard: How much guidance did you have during that process?

Ashvin: It was pretty cool. It’s very similar to the YC model of how they work with startups. We had weekly coaching sessions with our coaches, and there were lots of workshops along the way on design thinking and all these things you need to know—basically getting that foundation right for building a tech company.

Finding validation and refining the initial idea

We’ve spoken a little bit, Ashvin, about a bunch of stuff so far—funding, the initial inspiration. Let’s talk about finding and validating. You’ve been speaking about this a little bit earlier on. Identifying and achieving market fit is crucial for your business. And from what I’m hearing, a little bird has told me that you are a creator yourself, as are your other co-founders. I guess that helps you make sure you’ve got the right target audience because you were just saying, "If we would use it, then other people would possibly use it," is the thinking I’m guessing behind it. Once you’ve got that idea and know that you’re going to use it, how did you then refine it to make sure that not just you three or you four would use it, but other creators would use it as well?

Ashvin: Fun story. I think in the initial days, right when we were planning out our product as well, we actually launched an initial product that was more of a chat interface. We’ve since pivoted quite a lot, but we actually launched the chat interface and then tried using it ourselves. As we were using it, we realized, "Hey, this actually doesn’t make sense for our workflows, and it’s not actually what we needed." So it wasn’t so much of a "Let’s design the ultimate tool and then just build it." It was more about, even though we knew what problems we were facing, when we built out our initial product, we ourselves didn’t feel like it was valuable. So it was actually a series of iterations all the way until the last three months since our initial product launch. Every week, we’ve been pushing out updates, changing entire UIs, and changing entire systems just to try and find a product that even we ourselves would use every day. I think it was after one month of iterations, after interviewing lots of customers, that we finally found something that we’re using now in our personal branding journey as well. That journey has been fun. After that, it was about just creating content on our own personal brands. We were posting on LinkedIn, TikTok, Instagram, and basically just spreading the word. As a creator, I’d say, "Hey, I’ve built this tool. Would you guys like to try it out?"

How did you get your first paying customers?

Richard: On the back of that, then—and a question that was asked during the break and stolen by my producer—once you’ve created this product, how did you then attract your first real paying customer?

Ashvin: Basically, until today, we haven’t spent any money on marketing or advertising. It’s been fully organic through our own personal brands as well. Our first thousand users came in through us creating content, and a couple converted to paid users as well. It was actually a huge validation because, at the end of the day, we are constantly iterating through the product and trying to build something that people find really valuable.

Richard: From what I understand, it’s a subscription model. What are your recurring customers like? What’s the retention like?

Ashvin: We’ve only had paid users for the last month now. The product’s three months old, so I can’t say much on retention yet.

What's a typical day at Cleve like?

Richard: Fair enough. We’ll ask you again in a few months' time. Let’s get behind the scenes a little bit, just before we go into the news. What is a typical day at Cleve like right now?

Ashvin: I guess it’s a lot of figuring out the biggest leaks in our bucket. For example, if there’s a giant hole where our users aren’t finishing onboarding, we have to fix that first. Or maybe we have a low reactivation rate, so people use it for the first time and then don’t use it again the next week. So, we fix the email sequences, fix the onboarding, and ideate different product ideas, then build them. Day to day, it’s a lot of that product development process. I love this early stage of the company where we can change so much and get away with it. Customers are like, "Yeah, just play around with it. Experiment."

Richard: You were saying about the onboarding process, right? Is there a particular point at which people regularly stop onboarding? Do they get to a certain point and go, "Oh," and have you identified that area? What is it?

Ashvin: Honestly, our current onboarding process will be changed in like a week or so. It’s really long—it’s 15 steps right now. We’re cutting that down to three. We needed to have a really long process initially because the time that the model takes to process things was really long. So, instead of having them wait on a really long loading screen, we temporarily added a bunch of screens for them to go through so they don’t have to wait that long. But yeah, it’s lots of little things like this that were very hacky and we just had to put together.

How did you go about market research?

Richard: What kind of market research were you doing before you started building this product and as you were building it? There are similar things out there already. What kind of research were you doing?

Ashvin: It was largely focused on the personal branding space—what existing creators or personal brands or business leaders are using right now and how they’re finding it. There were a lot of customer interviews, really understanding how they’re finding ChatGPT, Gemini, etc., as well as more niche tools like Jasper and Copy.ai. What problems are they facing, and how can we build something better?

Will all content sound boring once everyone's using AI to create content?

Richard: Is there a worry that in the future, when tools like this are readily available and people are using them over and over again, content that appears online will start to feel a little bit homogenous?

Ashvin: Yeah, I think we’re already starting to see that, especially with more and more people using ChatGPT. The core philosophy that we believe in at Cleve, and even our core design philosophy, is that AI content doesn’t work if it’s just fully generated. The goal of AI content tools really needs to be taking your actual experiences, your actual ideas, and organizing them in a way that’s optimized for distribution. Creators, personal brands, and business influencers should be focused on their message and their story. AI can just help polish it and make it better—adding a better hook, making it more readable, so it gets the discoverability and reach it deserves. As long as people and their stories are at the focus, it’s going to be fine.

Richard: I’m interested to see how this works. I’m going to have a proper, in-depth play with it a little bit later on.

How do you differentiate from other platforms, especially in the AI space which can be highly competitive?

Let’s talk a little bit about competition. We briefly touched on it. This personal branding and tech space is highly competitive, with a variety of mediums available today. How do you differentiate yourself from other similar platforms? What would you say is your unique value proposition?

Ashvin: In our case, the real opportunity, or rather the gap we’ve seen in the market, is that existing tools lack context, tone, and personality. Hence, the content generated, even if you seed it with information, requires a whole bunch of prompt engineering to get anywhere with it. We basically try to make that whole workflow as simple as possible. I think that’s where we haven’t seen many players operating. But that said, my opinion on competition is that yes, although there are a lot of big AI tools out there, how many of them have actually made it into the workflows of business leaders or creators on a day-to-day basis? As we were talking to more business leaders and creators, they are using AI, but not on a very regular basis—they don’t have a platform they use daily. Competition has two sides: the number of players in the market (probably more than 10,000 globally), and the users—how many tools are they considering using? We feel like in the AI space right now, there’s still a big opportunity because many people haven’t adopted these technologies as they haven’t been delivered in a form that makes sense to their workflows.

Where are your users from?

Richard: You just mentioned that the majority of your users are not from Malaysia. Where are they from?

Ashvin: Singapore, the U.S., Europe, and Australia.

Richard: Where do you see your biggest market developing?

Ashvin: Most likely, it’s going to be in Western countries where personal branding is more common. People in those areas are also naturally more outgoing—they value their personal brand. We do see users in those areas, and also in Singapore, where they’re more vocal. But honestly, I personally feel like we should be more proud and talk louder about ourselves, even in Southeast Asia and in Malaysia. Early adopters are probably going to be in the Western markets, but I do hope to see more Malaysians because we have so many great stories here—so many great entrepreneurs as well. I’d like to see more Malaysians sharing their stories.

What's your revenue model?

Richard: Before we wrap up, let’s talk about your revenue model. I know it’s a subscription model. How do you generate revenue? After the revenue model, what have been the most effective revenue streams so far for your business? I’m guessing you played around with a few ideas first about how much you should and shouldn’t charge. How did you manage to land on the amount you’re on right now?

Ashvin: Basically, it’s freemium. We have a free tier where people can use the note-taking capabilities and stuff for free. The paid tier is $18 a month right now. It’s an experiment. With startups, it’s always going to be "just throw a number out, see how it sticks." How much do people value your product in their workflows? The $18 price point— we’ve seen a bunch of users sign up already, so that’s been great and has given us some early validation of the value we’re creating for users. Yeah, that’s basically it. Just throw a number out, see if it sticks.

Richard: And so far, it’s sticking?

Ashvin: So far, it’s sticking. We’ll probably play around with that in the future.

What was the biggest obstacle so far in the journey?

Richard: My final question for you—no, sorry, before I get to my final question. In terms of obstacles that you’ve faced so far, what do you think was the biggest? You’re a new company, and not every new company is easy to get through and easy to start. What’s the biggest hurdle you’ve had so far?

Ashvin: For us, it’s been tricky figuring things out in this space. How do you really build a tool that’s sufficiently differentiated and creates significant value for the end users from scratch? There’s no benchmark; there’s no competitor that we can actually look at because we’re designing our UX from the ground up. We’re going from first principles, and that design process can be really tricky. We’re testing out a whole bunch of different flows and seeing what makes sense. Just going through the iterations and having to kill our own darlings as well—when we’ve built a UI that we feel would be great, we test it out with users and nobody knows how to use it. We had that for probably the first two months, where every customer we’d onboard had no clue how to actually get value out of the product, even though it seemed simple enough to us. So, a lot of playing around with product design—finding a lot of emails from Ashvin asking, "How do I use this?"

Richard: Did you have any of those at the beginning?

Ashvin: It wasn’t even emails at the beginning. We would go to the customers, sit down with them, and say, "Hey, try out the tool. I’m just going to watch you." We would see how they would be clicking on a whole bunch of buttons that didn’t exist, or they couldn’t find certain buttons. Because we’re building from first principles, we’re not copying someone else’s UI. So, we made a lot of mistakes along the way, then fixed them. I guess now it’s a bit better.

Richard: I’m guessing that’s very useful for your feedback because you’re like, "Actually, I need this." How quickly are you able to take that advice or criticism and implement it into your model?

Ashvin: That’s the beauty of having a three-person team—literally the next couple of days. Sometimes, if it’s a simple fix, it’s within the same day. Our CTO Rashad’s just building and pushing that to production straight away.

How much sleep does Rashad get?

Richard: How much sleep does Rashad get?

Ashvin: That’s a question for him. I don’t want to say it on air.

Richard: But I think you guys are hardworking.

Ashvin: Yeah.

What's next for Cleve?

Richard: Okay, final question for you then. As you guys continue to grow, what are your plans for expansion? Maybe new features, maybe collaborations that you have in mind that you’re looking at. Can you talk to me about any of those?

Ashvin: So, we’re always exploring. On the product side, we’re always launching new things. The last couple of weeks, users have been coming to me every week being like, "Hey, what does this new thing do?" Because every weekend, we’re releasing and testing something new with users. In terms of expansion, it’s continuing to double down. So far, our only marketing channel has been our personal brands. We’ve just been posting content on LinkedIn and TikTok, and that’s actually been getting a lot of early users, which has been really great. But yeah, in the future, we’re looking at collaborating with key influencers in different spaces, particularly in productivity and stuff.

Richard: Radio interviews?

Ashvin: Yeah, perhaps. And we’re really doubling down, especially in the U.S. and overseas markets.

Richard: Ashvin, thank you very much for spending time with me this morning.

Ashvin: Absolutely. Thank you so much. It’s been a pleasure.

Richard: Folks, I’ve been speaking with Ashvin Praveen, the co-founder and CEO of Cleve.ai. Their local startup is aiming to change the personal branding landscape with their AI-powered content creation workspace. You have been listening to Open for Business, the show that dives deep into the journeys and ventures of entrepreneurs and business leaders. If you missed any part of this conversation, you can catch the podcast on our website at bfm.my or download the BFM app. You can also find our shows on Spotify, Apple Podcasts, and any other podcast players. Just search for Open for Business. It’s me, Rich Bradbury. Keep it here on BFM 89.9, The Business Station.